Gabriel Adamante

Token Split Tool - Easily Split text based on token count

Mon, 31 Jul 2023 13:10:13 GMT

I've built this because I kept running into context size length limitations on LLMs.

It enables you to split large bodies of text into smaller tokens based on a defined character count (assuming 1 token = 4 chars).

Enter your text and desired token size below and get an output.

Output:

Check out my new project to track prices on your favorite websites.

NFe Analytics: analyze data from NFe XML files, free and easy

Tue, 09 May 2023 22:44:43 GMT

This tool was built to help small business owners who want to understand more about their sales - revenue over time, top customers, and best-selling products.

To use it, just go to nfeanalytics.adamante.com.br and upload your XML files.

Why build NFe Analytics?

The idea behind NFe Analytics is to simplify the process of analyzing NFe invoices and managing revenue for small business owners who don't have the time or resources to deal with complex ERPs. NFe Analytics lets you manage your revenue with ease, without having to deal with complicated systems.

How NFe Analytics works

NFe Analytics analyzes invoice files without uploading or sharing your data, ensuring complete security and privacy.

With simple controls, you can filter the data by date, product, or customer, and the application will display total sales, sales by product, and sales by customer in a clear and concise way.

Also, since there is no server-side code or third-party services involved, NFe Analytics is completely secure and focused on user privacy.

Using NFe Analytics

To start using NFe Analytics, just visit the hosted version at nfeanalytics.adamante.com.br.

All data processing happens on your device, ensuring the same functionality and privacy as the local version. If you want to host it on your own server, or run it locally, here is the github address.

Finding the ideal apartment (Like a Hacker)

Sat, 21 Nov 2020 21:18:19 GMT

2018, the year I decided to head to Europe to open an açaí franchise (which didn't work out), was also the beginning of my experience with remote work and "digital nomadism".

While things weren't coming together to get the franchise off the ground, I kept working as a Digital Marketing freelancer (while learning to code in my free time).

At the end of the year, when I came back from Europe to Brazil, I turned that freelance gig into my "official" bread and butter with Traktor. I came back here and started traveling, spending a bit of time in each place.

And since then, I haven't stayed in any one home for more than three months straight.

This life has plenty of upsides, and I'm fully aware that being able to live it is a huge privilege.

But there comes a point when it gets tiring. At least for me, it did.

I miss simple things, like not having to worry about a place to keep my books and records.

Being able to keep a fixed routine. (I couldn't pull that off while traveling). Being able to scale.

Every time I change cities, I have to start everything from scratch.

All of that, plus other personal factors, led me to a decision: I'm going to stop being a digital nomad!

It's not a final decision, I might change my mind. But I want to spend at least a year in the same place and see if I feel better that way, putting down roots in one city.

Which brings us to the main topic of this article: finding the ideal apartment after so much time hopping from place to place.

Like a hacker.

Hackerman hunting for apartments

#1 - Defining the specs

I've spent plenty of time in flats and one-bedroom apartments over the last few years.

This time, I want a bigger apartment, with plenty of room for a guest bedroom and an office.

Furnishing a rented apartment is also something I'm not too inclined to do, so it has to come with at least the basics, like wardrobes and a kitchen.

I've also discovered that I love taking care of plants, and space for a little vegetable garden (even a vertical one) is very welcome - which means I have to worry about sunlight too.

I also want two bathrooms and a gas-heated shower, because I'm in Curitiba and it's cold as hell here.

Beyond that, I don't want an apartment waaaay out in the sticks - but I'm not dead set on being super central either. So I'm filtering for the following neighborhoods: Portão, Vila Izabel, Água Verde, Batel, Campina do Siqueira, Bigorrilho, Alto da Glória, Alto da XV, Centro, Centro Cívico, Rebouças.

Finally, one of the most important things: I'd like to pay up to R$2000 a month, rent and building fees combined. I could stretch that budget a little if I found the perfect apartment. But I'd rather not.

So, my specs are the following:

2 or 3 bedrooms
2 bathrooms
gas-heated shower
more than 50m² of usable area
3-4 hours of sunlight
furnished
in one of the neighborhoods listed above
up to R$2000 in total cost

So let the search begin!

#2 - Searching the obvious places

Like anyone would, the first thing I did when I decided I wanted an apartment was search Google for: Apartment Rentals in Curitiba.

If there's a bloodthirsty SEO market out there, it's this one. I know I looked at all the biggest portals (Imovelweb, Vivareal, Zap imoveis, Quinto Andar);

I had a lot of trouble properly filtering listings on those platforms. I figured that would've been solved by now.

A special shout-out for horrible usability goes to Zap Imóveis, which asks you to share your location on EVERY page. NO MEANS NO, ZAP.

Anyway, I needed to evaluate the apartments available on those platforms.

Using the filters (keeping in mind that the furniture filter isn't reliable, and neither is the gas shower one), I got the following number of apartments on each platform:

Imovelweb: 130 apartments
Vivareal: 73 apartments
Zap: 46 apartments
QuintoAndar: 4 apartments

Among those, QuintoAndar has the best listings, but the highest prices. To give you an idea, if I raise the budget to R$3000, there are 18 apartments. I guess that's justified by the perks of renting through them.

Now, to evaluate which platform is the best, I'm going to need a little trick 🤠. Scrapers.

Extracting the data from the platforms

Using Simplescraper.io I quickly extracted the data from Imovelweb, Vivareal and Zap to run a quick analysis of the kind of listings I'd find there.

It looks like Imovelweb lists bigger, more expensive places - even its average total price blew past my budget. Vivareal landed on a decent average, and Zap seems to have apartments that are needlessly expensive for their square footage.

I looked through all the options on those three sites and found some interesting ones... But nothing I fell in love with. All the really good places were expensive.

But one thing really got me suspicious...

I found almost no listings from the small local real estate agencies. And they usually have nice places listed.

#3 - Striking Gold

When I first moved to Curitiba, in 2016, I rented a place through a small agency, Firenze.

That was the first place I went looking.

And... right off the bat, I found 23 listings matching the specs I wanted!

via GIPHY

Great. Now, I just need to find a way to list every local real estate agency, and then their listings...

And I'll find the perfect apartment at one of them.

Super simple, right?

Finding every local real estate agency

For that, I went to the place where you can find almost every small business. Google. More specifically, Google Maps.

And sure enough, in the areas where I'm looking, I found at least 200 real estate agencies.

Now... How do I aggregate all this information the easy way?

Since I'm not crazy enough to build a Google Maps scraper on my own, I used Phantombuster - Google Maps Search Export.

Easy peasy. 200 agencies, sorted by neighborhood, with websites, in a spreadsheet.

Now, I need to process this information.

That's going to be a real challenge, since every site will be unique, and I have a feeling a lot of them will be badly built.

Since the process of choosing an apartment also has a lot to do with an incalculable qualitative factor - how much you like it - it's a bit hard to automate this part.

But there's no decision a good Excel spreadsheet can't help us make!

I manually went through the websites of 28 agencies and manually rated the listings I liked most. It took a few hours, but it looks like I found four places that check every box and are within budget!

Success!

Wrapping up the saga

I still haven't settled on one of these apartments. I'm going to ask a few friends for their opinions, and after that, contact the agencies.
See if anyone can offer a special deal too.

I learned a few things along the way:

The best value-for-money apartments are the hardest to find (obviously!)
Real estate agencies work as a network - several of them have the same listings on more than one site.
The real estate market is nowhere near offering a pleasant digital experience yet. Over the day I spent doing this, I managed to think of 2 or 3 business models that would make sense in this space.

Thanks for reading this far, and if you have any suggestions, send me a hello! :)

Histórias de Fim do Mundo (End of the World Stories)

Sat, 21 Nov 2020 10:47:36 GMT

The short stories gathered in 'Histórias de Fim do Mundo' are engaging and unpredictable. Blending ironic humor with subtle reflections about us and our surroundings, the fun and lightness of the reading are never set aside, held together by an atmosphere of magical realism that draws the reader in from the very first line and seeks to explore the possibilities of subverting the everyday, turning the banal and routine into the surreal and thrilling.

This is my first book.

It's a book of short magical realism stories. My biggest inspirations were Julio Cortázar, Jorge Luis Borges, and Gabriel García Márquez.

It was a simple book, but I dare say it's a good start. The illustrations are all by Lucas Costa Rodrigues, who did a beautiful job.

You can buy the book on Amazon, or download a PDF that I'm making available for free. (Download the PDF and then tell me what you think.)

The 6+1 Copywriting Formula: (sell, don't push)

Fri, 15 May 2020 12:46:30 GMT

The 6+1 formula is a copywriting model for texts with a bit more nuance than AIDA. It works to generate Attention, Interest, Desire, and Action all the same - but it does so in a more detailed way.

Who this formula is most useful for:

Anyone who is a beginner at copywriting and needs a more detailed formula to write with
Anyone who feels embarrassed about being too salesy and pushing their product/service onto people
Anyone who offers high-value services tied to their reputation: consulting of any kind, independent professionals, and the like.

Sound like you? Then let's get to the formula:

SUCCESS!

Like any other formula, it's basically a sequence of arguments used to get people to buy a service or product.

This is a slightly less "salesy" formula than others, with room for nuance when you're writing.

stolen (shamelessly!)Danny InyHere's the original.

The steps are as follows:

Step 1: Context

The first thing we should do, before calling attention to anything, is establish context.

We're going to answer the question: "Who are you, and why are you talking to me?"

Here you'd introduce yourself quickly and throw in some trust elements, maybe a credential (specialization), or a client who works with you.

Step 2: Attention

Once we've established context, we can grab people's attention.

There are lots of ways to grab attention. One of them is with a shocking statement, (Did you know that x% of people have problems with xxxx?).

I bet you already have some arsenal of interesting arguments for this, right?

Step 3: Desire

Whoever is watching the video already knows who we are and why we're relevant.

We already have their attention, and maybe they're aware of some problem.

Now we have to make them feel a very strong desire for what we want to offer

Here we paint the picture of how our product or service positively impacts that person's life.

We list how our service improved attribute X in person Y's life, we make promises (which will be kept!!), and we establish a desire to achieve that same result.

Step 4: The difference

Now we start moving toward getting people to make some decision - click a button, send a message, book an appointment - whatever it is.

We do this by communicating the difference between what will happen if they stand still, doing nothing, and what will happen if they seize the opportunity we're offering.

This is called establishing the difference.

The best way to do it is to ask: what if nothing changes? What will that mean for you?

Step 5: The Solution

We can't leave people not knowing what to do!

Once we've established the difference between taking action and standing still, we present the solution to their problems.

It's important to remember that we should talk about the solution, and talk about it enough for the person to understand that it works. But don't go on and on about it.

Step 6: Call to Action

Now, we need to establish one single action we want the audience to take as soon as they finish consuming the content.

Not a series of options (call, send an email, follow me on twitter).

Just one action.

And then we explicitly ask the audience to make that decision.

+ 1 step: Credibility

We want to keep our credibility.

That means being empathetic, showing you understand the situation deeply, and proving you can help them.

Our goal is to weave credibility into every step of the text/video, so that it's built slowly, and by the end it's fully established.

Book Summary: The Power of Thinking Outside the Box: The Lessons of 8 Executives Who Overcame Seemingly Unsolvable Problems

Tue, 28 Jan 2020 14:04:55 GMT

The book is a case study of 8 CEOs of public companies who achieved extraordinary returns in their companies' stock value, and isolates the factors that led them to such success.

Book: The Power of Thinking Outside the Box: The Lessons of 8 Executives Who Overcame Seemingly Unsolvable Problems

Author: William N. Thorndike Jr.

Rating: ⁸⁄₁₀

Date Read: January-2019

Description:

Really good! The author basically looks for extraordinary outliers relative to the market and over long periods of time, and tries to understand what points they had in common.

He managed to find several common traits, but here are some of the ones that caught my attention the most:

they were all people who avoided publicity in order to focus on their work.
extremely decentralized operations, extremely centralized capital allocation
The CEOs were the ones deciding capital allocation, and they never outsourced this task;
They always determined the minimum acceptable rate of return before making an investment / acquisition
They were great at calculating risks and returns
Instead of issuing stock, they bought it back (which was extremely unconventional)
They were all extremely careful with taxes, and always had great tax-specialist accountants to help them
Greedy when everyone was careful, careful when everyone was greedy.
They didn't hesitate to exit investments that were losing money, and to sell businesses when they were highly valued.

These were some of the CEOs' traits. There's a lot more! I recommend reading the book.

How to run a Product/Market Fit survey (the Sean Ellis Survey)

Thu, 16 Jan 2020 18:59:50 GMT

Summary: how to apply the Sean Ellis survey to assess your product's Product/Market fit.

Sean Ellis, author of Hacking Growth, described in his book a simple survey to find out whether your product has reached market fit and is ready to start an accelerated growth curve.

Basically, the survey starts with the following question:

How disappointed would you be if this product stopped working tomorrow?

Very disappointed
Somewhat disappointed
It wouldn't affect me much (I don't think it's all that useful)
I don't use the product anymore, so it makes no difference.

If more than 40% of the answers are "very disappointed", then you have Product/Market Fit;

Results lower than that can mean either that your onboarding is lacking (people don't know how to use the product), the product is lacking (the product doesn't meet expectations yet), or you have the wrong audience (we attracted the wrong kind of person to use the product).

In these cases, you need to understand where the problem is and fix it.

The questions that help answer these doubts are the following:

What would you use as an alternative to our product if it were no longer available?

I probably wouldn't use an alternative
I would use: ____

What is the biggest benefit you get from using our product?
Have you recommended our product to anyone?
If so, how did you describe it?
What kind of person do you think would benefit most from our product?
How can we make our product better so you're happier with it?

How to implement this survey:

The best way to run the survey is using Hotjar and its Polls feature, which is free for up to 2 thousand pageviews per day.

You can see their tutorial on how to do that here.

If you don't have a website, or want to send the survey by email, use Google Forms or Typeform to do it.

Good luck!!

How to Learn Growth Marketing

Mon, 11 Nov 2019 20:30:07 GMT

The goal of this guide is to create a step-by-step path to learn Growth Marketing and make life easier for anyone just getting started.

My motivation for writing it is simple: I had a really hard time finding high-quality knowledge when I was starting out in the field.

There's a lot of shallow material and clickbait out there.

Let me be clear: this is a curation of content that I consider high quality, with a clear order for your studies. It's not an exhaustive growth training program, but it gives you a solid base to start in the field.

But...

What is Growth Marketing? Or Hacking?

Growth Marketing, growth hacking or performance marketing is maximizing a company's revenue through the use of data.

Growth Hacking is just a buzzword used to refer to the same thing.

This is the definition from @julian, and it's the one I believe is most on point.

The difference between growth marketing and traditional marketing lies precisely in the precision with which we can measure our results.

Measuring the revenue lift a company gets from billboards is... an extremely difficult task.

Measuring the results of a Google ad may not be easy (Depending on how messy things are), but it's doable.

Another important distinction: growth marketing focuses on increasing a company's revenue, not on making people know your brand (Brand Awareness).

How to learn Growth Marketing?

Basically, to learn Growth, you have to understand the foundations of digital marketing and the foundations of the data analysis and experimentation process.

So you need to know how to operate a marketing funnel (acquisition, conversion, engagement, revenue and referral), and be able to experiment with it consistently.

All of this gets much easier if you're a technical person, but that's not required.

There are countless tools that make your life easier on that front.

Growth Team vs Growth Professional

A lot of people say it's impossible for there to be a growth professional, or a growth hacker 🤢, who concentrates all the skills needed to run this process on their own.

I disagree. @ryanckulp (FOMO) is a living example of someone who does the whole process, start to finish; @levelsio is another;

I'm sure you'll find endless examples of people who bring together enough skills to build and grow a business on their own, but that's not necessarily how it has to be.

With a team, things get simpler, since you can distribute the process and make it more accessible.

What matters is that it contains all the essential components, which translate into key skills:

User acquisition and conversion
Product development
Continuous experimentation

This guide sets out to cover User acquisition and conversion, and Continuous experimentation.

Learning about user acquisition and conversion

These are the skills you develop over a digital marketing career.

User Acquisition

SEO, SEM, Social Media, Influencer Marketing, Viral Marketing, Social Ads, among many others, are examples of user acquisition channels.

When the book Traction was written, the authors listed 19 traction channels for acquiring customers.

Today, some people already talk about 21 channels.

My recommendation is the following:

Get to know all the channels. You can do that by reading the book Traction, or read this post where Vinicius Camilo summarizes that information.
Learn to operate, at least at a basic level, the paid channels on Google and Facebook
To learn the others, experiment in an orderly fashion, using the Bull's Eye Framework, with the other channels.
You'll have more skill/affinity executing certain specific channels. Specialize in them.

Learning to operate paid channels is important because they tend to be the easiest to test, and they also validate other channels.

If Google Ads works for you, SEO should work too.

If Instagram Ads work, both Social Media Marketing and Influencers should work as well.

And you can validate these channels in a week or two (for most businesses).

Here's a list of the available channels:

(If you have a good intro link for any of these channels, send me an @ on Twitter or an email.)

Influencers
Public Relations
Unconventional Public Relations
Google Ads
Social media ads
Offline Ads
SEO
Content Marketing
Engineering as Marketing
Partnerships
Sales
Affiliate Programs and Channels
Other Platforms
Fairs and Trade Shows
Offline Events
Speaking Engagements
Community Building
Email marketing
Viral Marketing

User Conversion:

This tends to be the first bottleneck for anyone learning marketing.

We often focus a lot on how to operate a channel, but forget a few things that apply to all of them: persuasive language, clarity about the next step, and sales arguments;

That's why these skills are essential for a sales professional/team.

Without them, you won't be able to generate leads to make your sales, let alone move those leads down the funnel.

So the next step is learning how to

Be a unicorn: pick your Marketing Stack

Now that you have a sense of the whole process, you can choose a marketing stack to specialize in.

A programmer refers to their "stack" basically as the universe of programming languages/skills they operate fluently and get good results with.

Bringing that analogy to Marketing, we can think of a marketer's "Stack" as the universe of channels they specialize in and are most proficient at.

For example:

Paid Media
Landing Page Design and Copywriting
Email marketing

This Marketing "stack" guarantees you can run the whole marketing process in a specific way:

attract users using paid media,
convert users using copywriting and well-built landing pages,
make a sale (another conversion) using your email marketing skills.

Another possible stack:

SEO
Content Marketing
Email Marketing

With this Stack, you specialize in attracting users organically to make a sale.

There are countless ways to combine these skills and the operation of the 19 traction channels.

The more unique your stack, the more valuable your marketing skills will be.

Up to the point where you become a unicorn. A professional so unique and valuable they're almost impossible to find!

Think about it:

how many professionals specializing in paid media are there in Brazil? 20 thousand?
how many paid media specialists are also great landing page designers? 1 thousand?
how many paid media specialists are great landing page designers and know how to build an email funnel that generates sales? 100?

The idea behind this line of thinking is the following:

Past a certain level of specialization, your marginal optimization gains shrink. Returns are diminishing.

For example:

There comes a point, when optimizing an ad account where all the best practices are in place, where you've already hit 95% "account optimization*".

Tweaking the account structure, keywords, ads, and so on will only bring small gains.

You can go from 95% to 98%.

That corresponds to a 3.17% improvement in the final result.

But if your landing page is sitting at an "optimization index*" of 40%, for example.

Optimizing it to 80% will double your final result.

* Think of this as an arbitrary variable that defines a channel's optimization level. It's not meant to be taken literally.

That's why having at least 3 key skills multiplies your chances of getting good marketing results, and why you should pick a broader stack than usual.

Testing like a pro: continuous experimentation

To apply the tactics described here, you'll need to be really good at running tests.

The goal here is to run relevant tests at the highest possible speed, on a regular basis.

You need to know how to prioritize your tests effectively to pull this off. To do that, let's start by understanding the two types of test that exist:

Channel Tests (exploration)
A/B Tests (optimization)

Shall we start with the simpler one?

What is an A/B test?

An A/B test is an Optimization test;

It's used when you already have results from a landing page, media, creative, whatever - and you want to improve them.

To do that, you'll use the Split Testing method: you'll show the original version (control) and a new version of whatever is being tested to the same number of random visitors, and measure the impact on your goal.

Image stolen from Wikipedia

A good starting point to learn how to run an A/B test is this case study we did at Traktor.

One important thing: you should only start running A/B tests once you have baselines to test against.

In other words, there's no point A/B testing everything until you have significant traffic. It will simply take way too long, and you won't conclude anything.

To run an A/B test, you should do the following:

Run a diagnosis to identify the main improvement opportunities:

Sean Ellis's Product/Market Fit survey

Google Analytics Hotjar Mixpanel Heap

Prioritize one point to be specifically addressed by tests;
Define the goal of the test
Define the hypotheses to be tested
Define the method for tracking results
Collect and analyze results
Record the lessons learned

Know your way around spreadsheets (Google Sheets, Excel, Open Office, etc.). If you can pull data from a database with a bit of SQL, that's a nice plus;

You can use this A/B test tracking Spreadsheet to keep track of your test results.

What is a channel test?

A channel test is an exploration test;

You're exploring a new acquisition channel when you run this kind of test.

By definition, running this kind of test is a much more elastic task than an A/B test;

You don't have a defined baseline, you may have never operated that channel, and your results will vary a lot.

As you already know, there are 19 (or 21) channels to be tested, and to get through testing them, you have to prioritize.

Here's a Channel Test Prioritization and Tracking Spreadsheet using the Bull's Eye Framework;

The process goes like this:

Understand all the available channels and brainstorm how they could be used;
Prioritize the tests by Cost, Effort, Impact, and Lead Quality (see the spreadsheet);
Define the three channels to be tested over the next X period of time (it depends on the channel - ads usually show results within a month, content marketing and SEO take much longer)
Record the results, and interpret them;

Once you have enough information to decide whether a channel works, you can start focusing on it.

The Guide to Attribution Models in Google Analytics

Mon, 16 Sep 2019 20:37:56 GMT

How to use Google Analytics attribution models like a pro

Have you ever stopped to think about how many interactions a user has with your site before a conversion?

Yep. Some people take a pretty crazy path.

Assuming you got three clicks (like in the image) before a conversion, can you tell me which of the three channels was responsible for the sale?

The answer (almost always): it depends!!

Let's say you want to buy a hat!

You search for “Hats Curitiba” and get hit with an ad from Chapeleiro Maluco.

So you make your first click and discover the brand through a paid search. But you don't take action that day.

Two days later, you remember you need the hat and google “Chapeleiro Maluco”.

You make your second click, on an organic Google link.

And then someone calls you, you get distracted and close the page. No action taken on day two.

A day later, you type www.chapeleiromaluco.com straight into the browser bar and make a purchase through their ecommerce.

The conversion and your third and final click are recorded.

Which click was responsible for the sale?

Some people will say it was clearly paid search.

Some will say each one deserves a bit of the credit.

And some will say it was the last one.

The answer is: it depends on the attribution model.

Google Analytics' default model is last click, which is most relevant at the bottom of the funnel;

But there are other attribution models, like first click or linear.

Last Click Attribution Model

This is Google Analytics' default model, and the vast majority of people never even realize they're using it!

As the name suggests, the last click gets credited as responsible for the sale.

This model should be used to investigate the effectiveness of bottom-of-funnel actions, like email marketing, retargeting and branding campaigns, SMS blasts, and the like.

But it's often not the best way to look at your media investment, or your SEO, especially if your product / site has a longer consideration journey.

First Click Attribution Model

Another very simple model, it attributes the conversion value entirely to the first click.

Here we get a view of the top of our funnel.

You should use this model to investigate the performance of paid media campaigns, SEO and Content, Referrals, and so on.

The report gets especially powerful when you look at it across 90 days. (See below how to do it.)

Coming soon: Linear, Last Non-Direct Click, and Position-Based Attribution Models

How to use the Google Analytics Attribution Model Tool

First, go to the conversions report, attribution, and Model Comparison Tool

You'll be faced with this report:

In numbered order, the most important settings are the following:

1) Conversion Type

You often have several conversion types set up, and you don't want to analyze all of them. So it's very important to select which goal you're going to analyze:

I usually keep one or two selected:

If you select Google Ads, you'll be able to see the Drilldown of your campaigns.

2) Analysis Period

You can look up to 90 days into the past to analyze results for first-click campaigns, for example! I like to always keep this setting at 90 days.

3) Attribution Model Selector

Here you can select the attribution models to compare.

I suggest you start by selecting Last Click, First Click, and Last Non-Direct Click;

4) Secondary Dimension

We often forget about this option: here, you can select the report's primary dimension. The default is MCF Channels, but you can select Campaigns, Source / Medium, and even Landing Pages;

That way, you can explore the conversion path of an organic piece of content from your Blog by analyzing Landing Pages as the primary dimension;

5) Date Selector

Never forget to select the date!!

Remote Work: the good, the bad, and the ugly

Tue, 27 Aug 2019 15:04:31 GMT

It's been almost a year and a half since I started working remotely.

During that time, I've had wonderful experiences with remote work, and others not so much.

So I decided to write a bit about the pros and cons of working remotely: the good, the bad, the ugly!

I'll keep updating this as more insights come up ;)

Productivity

The Good

Working remotely means it's easier for me to spend long stretches of time focused, uninterrupted, in a state of deep work.

For tasks like development, writing - basically everything that demands my complex problem-solving skills - this is one of the most productive things there is.

The Bad

Sometimes, you don't need introspection. You need to go somewhere, talk to a person, and solve a problem.

In those cases, remote work can get in the way of things moving forward. Especially when you're an agency dealing with clients who don't understand this paradigm yet.

Although in an already-distributed organization it's easier to resolve this kind of situation with a call or a text message, it's not always like that! Even more so when you mix an in-person organization with a remote one - that can cause friction.

Freedom

The Good

For me, this is the essential point of remote work.

Alt forms of rich:

You can go to bed and wake up when you want to.

You can buy any book you want.

You have time to read those books.

You have time to exercise.

A short commute.

No dress code seven days a week.

Liberal use of the thermostat.
— Collaborative Fund (@collabfund) July 24, 2019

The freedom to work in the ways that make me most productive, and happiest, increases my quality of life very significantly.

These "alternative forms of wealth" are worth far more to me today than pure financial compensation.

The Bad

It's not necessarily bad, but freedom demands discipline. It's not always easy to wake up every day and do all the things you need to do, over a long period of time, sometimes with no one to hold you accountable.

The Good

Remote work gives you the chance to be closer to some people, like your family.

The possibility of a mother being able to give her kid attention and work without having to stick to strict schedules and commute all the time, for example.

But, usually, that's not how it goes.

The Bad

Sometimes, human contact does a lot of good. (Sometimes, hehehe).

A relaxed work environment, where you can socialize with people, have good lunches and things like that, makes a big difference.

The main problem with remote / nomad work is loneliness; I felt this quite a bit when I was in Valencia, where it took me a while to build a network and meet cool people.

What saved me over there was a good coworking space. Having a good place to work and meet cool people is really important in this process.

Now, we at Traktor are trying to bring the team together more often so everyone can get to know each other better, and also to create that relaxed work environment, strengthening the bonds between us!

Whatsapp Button Generator: put on your website in 2 minutes - WhatsWidget

Fri, 19 Jul 2019 22:07:10 GMT

Put a WhatsApp button on your website in 2 minutes: generate the button and paste the HTML code

Check out my new project to automatically screen resumes using AI: Magic People AI - Resume Scanner

1 WhatsApp Number 2 Widget Header

3 Message 4 Code:

❕Check out my new project to track prices in your favorite websites.

How to use:

Just paste the code somewhere you can put custom HTML/CSS.

It would be best to incorporate the CSS with other custom styles, but it’s not mandatory.

How does it work?

It’s just simple CSS using media queries to display two different links: 1 for WhatsApp Web, for larger screens, and other for the WhatsApp API, for smaller screens.

Wordpress:

I’ll make a plugin in a couple of weeks.

For now, you can paste it in the footer of your website:

Style > Editor

Go to the file footer.php and paste it in the beggining of the file.

Do not change anything else in this file. Do not delete anything if you don’t know what you are doing.

Don’t forget to update the file.

Aggregators and Platforms will swallow the world

Thu, 11 Apr 2019 12:56:00 GMT

What do Apple, Amazon, Alphabet (Google), Facebook, Microsoft, and Alibaba have in common? They are aggregators and platforms.

Of the 10 largest companies in the world, 6 are aggregators or platforms.

And it's quite likely that this number will be even higher in the future.

What is the one thing that Apple, Amazon, Alphabet (Google), Facebook, Microsoft, and Alibaba have in common?

Yes, they are all technology companies. But that's not the key trait that made them so gigantic.

They are aggregators and platforms.

Netflix has the same business model: aggregating movies and shows from various networks and studios (its suppliers) in one place.

The clever trick is that as soon as an aggregator has an interesting user base, its suppliers become more and more interested in joining the aggregator.

And the more suppliers an aggregator has, the more interesting it is to end users.

This creates a virtuous cycle that:

lowers the CAC (customer acquisition cost) for each new customer
lowers the operational cost of the business with each new customer

In other words: aggregators have a negative marginal cost to acquire a new customer.

In businesses that are not aggregators and platforms, the more your user base grows, the more expensive it gets to acquire a new customer: usually, the initial customer base has an ideal product-market fit.

As it grows, the further you drift from this audience, the more the value of your product decreases, and it soon becomes negative.

That's why in aggregator and platform business models, the winner takes all.

Competitors find it harder and harder to capture new users and get left behind.

The Complete Guide to Conversions in Google Analytics

Mon, 05 Nov 2018 12:30:00 GMT

A guide that goes from installing conversion tags using the main methods, all the way to complex topics like conversion attribution models.

I want you to leave this page knowing everything I know about Google Analytics conversions.

This is a complete guide, going from installing conversion tags using the main methods, all the way to complex topics like conversion attribution models.

If you already have conversions set up on your site and want to know how to use them like a pro, click here to jump to that part of the post.

I also threw in a bonus at the end, so you can use conversions at 101% efficiency.

What Conversions Are

Why are conversions important?

How to Install Conversion Tracking

How to pick the best method
Conversions via PageView
Conversions via Events
Sending events with Google Tag Manager
Sending events with WordPress Plugins
Sending events with JavaScript

3.Importing sales data into Google Analytics

Recording a User ID in Google Analytics
Sending data to Analytics through the Measurement Protocol]

4.How to use Goal reports like a pro

Why so much emphasis on conversions?

BECAUSE CONVERSIONS ARE THE MOST IMPORTANT PART OF WEB ANALYTICS

and also….

THE MOST EASILY IGNORED!

I've seen cases of people investing thousands of reais in AdWords without measuring conversions. 😠😠😠

If that's the plan, you might as well just burn your money.

If you're running any kind of online campaign, producing any kind of content… you need to measure your conversions.

Let me be even bolder: if you have a website,

you need to measure conversions.

If you know me in real life, just know this is the moment I'd be red in the face.. Worked up. Emphatic.

-(Me) MEASURE THE DAMN CONVERSIONS!

-(You) Ok, you lunatic..

-But, what are conversions?

What are Conversions?

A conversion is an action, on your site or not, that matters to you.

For me, on this site, it's signups to my email list:

If you liked this post, you're probably interested in the topics I write (and read) about.

I share a new post every Monday, along with interesting articles and books.

Give it a try and see if you like it:

I work with several companies at Traktor, and each one measures different conversions.

For Lead Generation businesses, common conversions are:

Leads signing up through a form
Phone Calls (You can measure calls like this)

If you work with e-commerce, you can measure conversions like:

Added an item to the cart
Applied x Discount code
Bought a product

Depending on the kind of funnel you have (high touch vs low touch), you can record sales conversions in SaaS and Lead Gen too.

You can also measure content conversions:

Lead magnet downloads (top-of-funnel conversions)
Email list signups
Video views on your site

In short: figure out which actions matter to you, and measure them as conversions

So far, I hope I've explained what conversions are. But I haven't given you a single reason to believe they're that important.

Why Conversions matter

Google Analytics loses half its usefulness if you don't measure conversions.

Without them, you don't get access to the Conversion reports. They tell you:

which channels drive the most direct sales
which channels drive the most indirect sales
how long someone takes to buy on your site
how many times a user visits your site before a sale

That alone is more than enough reason to care about them, but

If you're putting money into paid media

AdWords, Facebook, Instagram or Twitter

Conversions become 10x more important.

Because they're what feed those platforms' artificial intelligence algorithms,

finding your ideal customer at the lowest possible CPC.

Measuring conversions in these situations can make your ROI grow explosively.

To be clear: Google Analytics conversions are not the same thing as conversions in AdWords, Facebook and Twitter.

You can even integrate Google Ads and Google Analytics so they measure the same conversions.

The definition stays the same - a conversion is an action, on your site or not, that matters to you.

But Facebook and Twitter do this differently, with other trackers.

I plan to write everything I know about the subject for each of these platforms.

Until I get around to it, I'll leave you with the official guides from:

How to install Google Analytics Conversion Tracking

In the simplest terms possible, you should choose the best tracking method.

Then, you implement it using:

Google Tag Manager (my favorite)
WordPress Plugins
JavaScript
Importing sales data

How to pick the best tracking method:

It depends on your conversion.

Remember:

Uma conversão é uma ação, no seu site ou não, que é importante para você.

In this guide I'll cover the most common tracking methods.

If I missed something, send me an @ on Twitter or an email and let's talk.

Conversions via PageView:

This is the easiest method to implement, it fits most cases, and it's the only one you can set up entirely inside Analytics.

You should probably pick this method for cases like:

Leads signing up through a form
Lead magnet downloads (top-of-funnel conversions)
Email list signups

It works like this: when the user lands on a certain URL, a conversion is counted.

You could count a Lead conversion, for example, by redirecting your user to a page after they complete a form.

This conversion can be set up straight from Google Analytics, following these steps:

Step by step to Set Up a Pageview Conversion in Google Analytics

Go to Admin

In the third column, under View Settings, click Goals

Click new goal:

In Goal settings, you'll choose:

A goal template, or a custom goal.

If there's a template that describes your action exactly, pick it.

If not, go with custom.

Then, in Goal Description:

Pick an easy, descriptive name for your goal. In my case: Lead.

Choose the goal type as destination

In goal details:

Choose begins with and your page's URL, ONLY WHAT COMES AFTER THE SLASHES, with no trailing slash.

For example: this page is

https://www.adamante.com.br/code/conversao-google-analytics/

It would look like:

code/conversao-google-analytics

Don't fill in the optional values, and save. (Unless you've calculated the value of a Lead — in that case, put it in Value)

Done: you're converting

Conversions via Events:

You can use Analytics' Event tools to record conversions.

They're very useful for cases where you're not on a traditional web page.

It works with Apps and dynamic pages where pageviews are virtual, driven by JavaScript

(all the frameworks: jQuery, VueJs, React, Angular and EmberJs).

You can also use this to record video view conversions.

Or record engagement with some feature of your site: I use it here on my WhatsApp button Generator.

Events should be useful for cases like:

Added an item to the cart
Applied x Discount code
Bought a product
Video views on your site
Engaged with dynamic content

To do this you need to:

Set up an event in Analytics
Send the event.

(How to send events through: Google Tag Manager, WordPress Plugins, JavaScript)

Setting Up Event Goals in Google Analytics

You can follow the instructions on how to create a goal in Google Analytics up to the part where we select the goal type.

Select the Event goal type and configure it with something like:

Remember, you're the one who “makes up” the categories, actions and labels. We'll send them through custom events in just a bit. Don't worry too much about what you call them at first, just make sure it's something consistent and that it's recording all the information you need to know.

Sending Events with Google Tag Manager

Google Tag Manager is a great solution for managing your site's trackers.

You can also record events using it.

As an example, I'll explain how to record a button click as an event.

Going back to basic HTML, a button would be marked up like this:

<button id="clique-em-mim">Clique em mim</button>

If you use Bootstrap or some other CSS Framework, it could also be something like:

<a id="botao-clique-em-mim" class="btn btn-primary">Clique em mim</a>

To record an event with Google Tag Manager, you need to know the button's ID.

In both cases, the ID is “botao-clique-em-mim”.

To find out your button's ID, you can right-click the element and choose inspect element.

A sidebar will open, with a node highlighted in blue. Whatever is between ID=“” is your ID. In this case, the ID is botao-header.

With that in hand, let's head to Google Tag Manager.

If you haven't installed it yet, you can follow this Google tutorial that shows you how.

After that, we start everything by creating a trigger.

In the right-hand side menu, click Triggers, then New.

Give the Trigger a name, something like “Clicked the Click Me Button”.

Click the central block, Trigger Configuration, and select Click - All elements

In the central block select This Trigger fires on Some clicks

Select Click ID, equals, your ID.

Save everything and your trigger is ready :)

Now, let's create a Tag.

In the Side Menu, go to Tags and click new.

Give the Tag a name. For us: **Analytics Conversion - Click Me Button”

Choose “Google Analytics - Universal Analytics”

In the first option, set Event.

Set the Category, Action and Label that you chose in Google Analytics.

You can ignore the other fields, but don't forget to select the Google Analytics setting you created when you installed the tag manager.

Save and scroll down the configuration until you reach Trigger.

Select the Trigger we just created, and save.

Your event is created! Together with the analytics conversion setup, you should now be recording conversions.

Sending Events with WordPress Plugins

You can use WP Google Analytics Events for that.

It supports Scroll Events and Click Events - you can use the ID like I explained above to record a click event.

I really don't recommend recording events through a WordPress plugin. I know it always looks like the easiest solution, but with every plugin installed in WP, you end up sacrificing a bit of the platform's performance, on top of exposing yourself to security vulnerabilities.

The best solution ends up being Tag Manager, which makes it easy to centralize all your tags and conversion events.

Sending Events with JavaScript

If you prefer doing things through code and avoiding firing scripts whenever possible (like GTM), this is a great option.

Using the following code, you can send the event to analytics:

ga('send', {
  hitType: 'event',
  eventCategory: 'Engagement',
  eventAction: 'generate',
  eventLabel: 'WhatsApp Button'
});

You just have to choose the event's Category, action, and name.

For example, on this site's “Subscribe” button, I put the code in an HTML onclick tag:

<button type="submit" value="Se inscreva" onclick="ga('send', {hitType: 'event',eventCategory: 'Conversion',eventAction: 'signup',eventLabel: 'newsletter'});" ....>

If you inspect this button right here, you'll notice it:

If you liked this post, you're probably interested in the topics I write (and read) about.

I share a new post every Monday, along with interesting articles and books.

Give it a try and see if you like it:

Importing Sales data into Google Analytics

This is a more advanced type of conversion, meant for those who are scaling their business and chasing maximum performance.

It's one of those Web Analytics investments that can pay off in gold, especially if you use it together with Performance Marketing, sharing conversions and defining audiences in Google Ads.

To do this, you need to store the Google Analytics ID (cid or uid) in your CRM or in a database of your own.

Then, you feed Analytics with the sales data and, if possible, its monetary value.

You can do this in Google Ads by importing a CSV file, but in Analytics the implementation is a bit more complex.

We need to use the Measurement Protocol for that.

The basic process goes like this:

You record an Analytics user's ID in some system
You record the sale information through your own system
You send the data to Google Analytics through the Measurement Protocol

Note: Remember that you should never send identifiable information about your users to Analytics.

In other words: you can send transaction data, which products someone bought, how much they spent, etc.

But you can't send emails, phone numbers, names, etc..; No personal information should ever be sent.

Recording a user's ID in Analytics:

Analytics generates a unique ID for every user who visits your site.

This ID is stored in a cookie, and you can use a bit of Javascript to record this information in forms that capture your Leads' data and send it to a CRM (or Ecommerce system).

That way, you'd assign a Google Analytics ID to each customer in your CRM.

If you're using the latest version of Analytics, you can use this code to find the Client ID:

ga.getAll()[0].get('clientId');

How to integrate this with a form depends, but in general, you'll create a Hidden Input and fill it in using javascript. It would look something like this:

HTML:

<form>
<input type="text" style="display:none;" id="formClientID" value="">
</form>

Vanilla Javascript:

var formClientID = ga.getAll()[0].get('clientId');

document.getElementById("formClientID").value = formClientID;

Ideally, you'd save this script in an external file and defer its execution.

Now you should be recording Analytics data in your CRM.

The next step is…

Sending the conversion data to Analytics through the Measurement Protocol

The Google Analytics measurement protocol works like an API that lets you post data.

Let's say you're using a CRM like Pipedrive to manage your sales, and you want to send that data to Google Analytics.

You could build an automation using Zapier, so that whenever a deal is marked Won, you send its information to Analytics.

To send the information, you need to make an HTTP POST request to http://www.google-analytics.com/collect, with a few parameters in the payload:

v=1              // versão do measurement protocol.
&tid=UA-XXXXX-Y  // ID de rastreamento / ID da propriedade.
&cid=555         // O ID do cliente que você registrou 
&t=              // Tipo de interação. Pode ser pageview, evento, etc.

In our case, we're going to record a sale event. For that, you need the following parameters:

&t=event         // Tipo de Hit, evento. Não mude.
&ec=conversion   // Categoria do evento. É obrigatório.
&ea=purchase     // Ação do evento. Também é obrigatório.
&el=             // Nome do evento. Opcional.
&ev=300          // Valor do evento. Você usa isso para registrar dados monetários.

Zapier lets you run a Python script as soon as the Pipedrive trigger fires:

if input_data:
    cid = input_data['cid']             // Você guarda os dados que puxou do Zapier em variaveis.
    value = input_data['value']         // Lembre de NUNCA guardar dados pessoais no Analytics.

payload = {'v': '1', 'tid': 'UA-XXXXX-Y', 'cid': cid, 't': 'event', 'ec': 'conversion', 'ea': 'purchase','ev': value}
r = requests.post('http://www.google-analytics.com/collect', params=payload)

Done! You've recorded a sale event in Analytics using data from your CRM in a completely automated way ;)

How to use Goal Reports like a pro:

Some time after setting up conversions, we'll start having data to analyze. This is where things get fun :)

Goal reports give us insights into how our most valuable users interact with our site.

I won't touch the E-commerce reports here since they deserve an article of their own.

The first group of reports, Goals, gives you a more general view of the goals on your site - they're mostly basic reports that don't need much explanation.

The report I find most interesting in this group is Reverse Goal Path. (I'll leave the Funnel Report aside because for that kind of analysis you're better off using a dedicated tool, like Heap Analytics or MixPanel).

Reverse Goal Path

Here, you can visualize how people move through your site's pages before completing a goal.

That way, you can get an idea of which paths are most effective at converting a Lead, and discover the content consumption process on your site that gets someone interested in your product.

By combining this with the Segments tool, we can slice that same path by several dimensions, like Device Type (Mobile vs Desktop), Acquisition Channel, and so on.

Where number 1 is, you pick which goal to analyze. Where number two is, we pick a segment.

The first column is where the goal was completed - if you're using a conversion page, it will always be that one. The next columns are the previous pages the person went through. Where it says (entrance) is where the person entered the site.

Say you do a lot of content marketing and SEO — if you Segment (where the 2 is in the image) the report by Organic Traffic

And from there, understand which content discovery path on your site leads to the most conversions. With that, a CRO analysis on the key pages could boost your lead generation significantly.

Multi-Channel Funnels

I believe this one, along with attribution models, are my favorite GA reports.

It's natural for companies that sell and generate leads online to have complex marketing funnels with many channels and touchpoints involved.

One of Analytics' great powers is being able to aggregate all this complex information, pointing out where to focus your optimization and investment efforts.

Shall we get started?

All the reports in this group have the same options at the top:

The Conversion Type

By default, all conversions come selected. So pay attention — if you only want to see a Lead conversion, or a sale, make sure it's the only one selected.

The Lookback Window

This is the past time period you want to include in the report. You should adjust it according to your customer's buying journey. Usually, for B2B products, 90 days is the best option.

Multi-Channel Funnels Overview

There isn't much going on here… Except for this amazing Venn diagram of your Conversion Acquisition channels.

My suggestion here is to set a longer time period so you get more data volume.

You can see from the diagram above that even though the Organic channel matters more in conversion volume, Paid Search and Organic Search share only a small overlapping area.

That means my Paid Search investment is being directed at a different audience than Organic, and that we're generating Leads that wouldn't have been generated without that investment.

Assisted Conversions

An assisted conversion is counted when some dimension (pages, acquisition channels, ads, etc.) measured by Analytics is part of a user's conversion path, but not their final interaction

I wrote a dedicated post about this report, which you can find here.

Attribution Models

My favorite G.A report!

With it, you get a holistic view of your funnel, how long it takes to acquire a customer, and which channels are responsible for which stage of the process.

See how to use it here.

Learn to Use Markov Chains by Building a Lyrics Generator That Mixes Gospel and Brazilian Gangsta Funk

Mon, 29 Oct 2018 11:59:00 GMT

Using Python and a bit of creativity, you can generate some pretty funny lyrics.

A little nerd humor:

I was studying Markov Chains and how to use Python to work with them.

I ended up building a song lyrics generator.

Naturally, I decided to mix funk proibidão (Brazilian gangsta funk) with gospel music to see what comes out.

(Note: If you don't find this as funny as I do, you're probably a normal person and I just have a really weird sense of humor.)

This was one of the lyrics:

It's not very coherent (I couldn't get it to rhyme!), but it was great for laughing and for learning Markov Chains.

How I did it: I built a scraper that downloaded the lyrics from Vagalume, then used a simple chain to generate the lyrics.

Then I found out there was a library that does it 10x better. (obviously!)/

If you enjoyed this post, you're probably interested in the topics I write (and read) about.

I share a new post every Monday, along with interesting articles and books.

Give it a try and see if you like it:

What is a Markov Chain?

A Markov Chain is a model that computes a sequence of events.

That sequence is computed in such a way that the probability of an event happening depends on the previous one.

This model is brilliant for working with text: it generates sentences by choosing words based on the previous ones, computing which words are most likely to come next.

Each event computes its probability separately

How to build a Markov Chain in Python

Now, on to our song lyrics generator.

The first part of the process involves having lots of lyrics in a text file, to train our chain on the probabilities of word sequences.

For that, I built a scraper for the Vagalume website, which downloads all of an artist's lyrics into a text file. I'll write a detailed post on how to do that soon. In the meantime, you can check out the final code.

The second part of the process is writing the Markov Chain itself:

We start by generating a random word.

After that, every other word is generated based on the probability of it following the previous one.

Shall we?

import random
import os

#Abrindo e lendo um arquivo de texto com as letras
lyrics = open("letras.txt","r")
lyrics = lyrics.read()

## Criando uma lista de palavras baseado no texto. 
## Aqui nós: substituímos as quebras de linha por espaços
## separamos o string pelos espaços, 
## usamos um filtro para remover valores vazios da lista

wordList = lyrics.replace('\n', ' ') 
wordList = wordList.split(' ') 
wordList = list(filter(None, wordList))

We now have a cleaned-up list with every word in our text file. The next step is turning it into a probability dictionary:

index = 1
chain = {} ## nosso dicionário
wordCount = 5 ## o número de palavras geradas

## um loop que busca todas as palavras na nossa lista, e coloca no dicionário uma nova chave com cada palavra, e seu valores, todas as palavras que as sucedem 
for word in wordList[index:]:
    key = wordList[index - 1]
    if key in chain:
        chain[key].append(word)
    else:
        chain[key] = [word]
    index += 1

Now, we start generating the lyrics:

#Uma palavra inicial aleatória
notCapitalized = True

while notCapitalized:
    firstWord = random.choice(list(chain.keys()))
    if firstWord[0].isupper():
        notCapitalized = False

We build a Loop that picks a random word based on the successors:

iterations = 0
while len(line.split(' ')) < wordCount:
    capitalized = True
    while capitalized:
        iterations += 1
        nextWord = random.choice(chain[firstWord])
        if nextWord[0].islower():
            capitalized = False
        if iterations > 15:
            capitalized = False
            iterations = 0
    firstWord = nextWord
    line += ' ' + nextWord

Tying it all together in a function:

def generateLine():
    wordCount = 5 ## o número de palavras geradas
    notCapitalized = True

    while notCapitalized:
        firstWord = random.choice(list(chain.keys()))
        if firstWord[0].isupper():
            notCapitalized = False

    line = firstWord
    iterations = 0

    while len(line.split(' ')) < wordCount:
        capitalized = True
        while capitalized:
            iterations += 1
            nextWord = random.choice(chain[firstWord])
            if nextWord[0].islower():
                capitalized = False
            if iterations > 15:
                capitalized = False
                iterations = 0
        firstWord = nextWord
        line += ' ' + nextWord
    return line

Now, we write a Loop that generates verses based on those lines:

index = 0
verseList = list()

while index < 12:
    verseList.append(generateLine())
    if index == 5:
        firstVerse = verseList
        verseList = list()
        print("\n")
    if index == 11:
        for y in firstVerse:
            print(y)
        print("\n")
        for y in verseList:
            print(y)
        print("\n")        
        for y in firstVerse:
            print(y)
        print("\n")
        for y in verseList:
            print(y)
    index += 1

Code

In the end, I found a library called pymarkovchain, which made the lyrics more coherent.

If you're only interested in using Markov chains to learn, it's much more productive to build one from scratch, but if you want to create something more coherent, it's worth taking a look at the existing libraries.

Markov Chain

## Esse método é baseado no código de Mehrab Jamee 
import random
import os

#Abrindo e lendo um arquivo de texto com as letras
lyrics = open("letras.txt","r")
lyrics = lyrics.read()

## Criando uma lista de palavras baseado no texto. 
## Aqui nós: substituímos as quebras de linha por espaços
## separamos o string pelos espaços, 
## usamos um filtro para remover valores vazios da lista

wordList = lyrics.replace('\n', ' ') 
wordList = wordList.split(' ') 
wordList = list(filter(None, wordList))

index = 1
chain = {} ## nosso dicionário
wordCount = 5 ## o número de palavras geradas

## um loop que busca todas as palavras na nossa lista, e coloca no dicionário uma nova chave com cada palavra, e seu valores, todas as palavras que as sucedem 
for word in wordList[index:]:
    key = wordList[index - 1]
    if key in chain:
        chain[key].append(word)
    else:
        chain[key] = [word]
    index += 1

## nossa função que encontra as palavras seguintes aleatoriamente, baseado no dicionário e nas anteriores 
def generateLine():
    wordCount = 5 ## o número de palavras geradas
    notCapitalized = True

    while notCapitalized:
        firstWord = random.choice(list(chain.keys()))
        if firstWord[0].isupper():
            notCapitalized = False

    line = firstWord
    iterations = 0

    while len(line.split(' ')) < wordCount:
        capitalized = True
        while capitalized:
            iterations += 1
            nextWord = random.choice(chain[firstWord])
            if nextWord[0].islower():
                capitalized = False
            if iterations > 15:
                capitalized = False
                iterations = 0
        firstWord = nextWord
        line += ' ' + nextWord
    return line


## a função que gera estrofes:
index = 0
verseList = list()

while index < 12:
    verseList.append(generateLine())
    if index == 5:
        firstVerse = verseList
        verseList = list()
        print("\n")
    if index == 11:
        for y in firstVerse:
            print(y)
        print("\n")
        for y in verseList:
            print(y)
        print("\n")        
        for y in firstVerse:
            print(y)
        print("\n")
        for y in verseList:
            print(y)
    index += 1

BeautifulSoup as a Scraper and pymarkovchain

```python
import os
from bs4 import BeautifulSoup
from pymarkovchain import MarkovChain
import urllib.request
```

## Vamos buscar o nome de todas as músicas do Mc Daleste

htmlArtista = urllib.request.urlopen("https://www.vagalume.com.br/mc-daleste/")
soup = BeautifulSoup(htmlArtista, 'html.parser')
nameBox = soup.find("ul", attrs={"class": "tracks"})
nameBox = nameBox.find_all("a")


## Colocando em uma lista somente as urls para facilitar o trabalho
nameList = list()
for i in nameBox:   
    nameList.append(i['href'])

## Agora, nós usamos a urllib para fazer o download de todas as páginas. (você também pode usar requests)
htmlList = list()
for name in nameList:
    htmlList.append(urllib.request.urlopen("https://www.vagalume.com.br" + name))    


## Aqui, extraímos letra das páginas 
songList = list()
for page in htmlList:
    soup = BeautifulSoup(page, 'html.parser')
    nameBox = soup.find("div", attrs={"itemprop": "description"})
    lines = str(name_box).replace("<br/>", "\n")
    lines = lines.replace('<div itemprop="description">', '')
    lines = lines.replace('</div>', '')
    songList.append(lines)

## e finalmente salvamos o resultado em um arquivo de texto, para facilitar a manipulação posterior
file = open("daleste.txt","w") 
for letra in song_list:
    file.write(letra)

## Agora começamos com a Markov Chain
mc = MarkovChain("./markov")

## eu juntei algumas letras diferentes usando o scraper de cima.

songlist = open("todas_letras_gospel.txt","r")
songlist2 = open("todas_letras.txt","r")
songlist3 = open("todas_letras_bk.txt","r")
songlist4 = open("mr-catra.txt","r")
songlist5 = open("daleste.txt","r")

## eu sei que isso é feio..

read = songlist.read()
read2 = songlist2.read()
read3 = songlist3.read()
read4 = songlist4.read()
read5 = songlist5.read()

read6 = read + read2 + read3 + read4 + read5

#geramos uma database usando todo esse texto que está junto:
mc.generateDatabase(read6)

# isso deve gerar uma estrofe - pra mim saiu: 'Ou me mama porque tu me abraça forte e não solte jamais' 
mc.generateString()

#aí eu faço esse loopzinho pra gerar uma letra completa:

i = 0
verseList = list()
while i < 12:
    verseList.append(mc.generateString())
    if i == 5:
        firstVerse = verseList
        verseList = list()
        print("\n")
    if i == 11:
        for y in firstVerse:
            print(y)
        print("\n")
        for y in verseList:
            print(y)
        print("\n")        
        for y in firstVerse:
            print(y)
        print("\n")
        for y in verseList:
            print(y)
    i = i + 1

One more bonus lyric for those who made it this far:

Meet Alpha-Beta: the Best Adwords Campaign Structure

Tue, 23 Oct 2018 11:57:00 GMT

If you advertise on Google Ads (the late Google Adwords), you need to pay attention to the most powerful - and basic - technique to lower your costs, increase conversions and CTR, and improve every single metric.

Whatever you call it - PPC or SEM - search advertising tools were optimized to generate as much revenue as possible… For the search engines.

Luckily for us, there are a few techniques that let the Advertiser squeeze out more efficiency.

This is the first article in a series on SEM where I'll go deep into Adwords and Bing Ads - how to get better results at a lower cost, how to automate your campaigns with JavaScript, and how to track conversions more efficiently.

This will be you after optimizing your campaigns.

Today we'll go deep into how to structure your Adwords account, and compare two methods that - in my experience - are the best ones for the job.

What are Campaign Structures?

Campaign structures are ways of organizing your Adwords account that let you control when and where your ads get shown.

Poorly structured accounts are inefficient and waste your money.

So when you structure your account the right way, you guarantee:

Confidence that you're advertising on keywords that actually make sense
Higher Quality Score on your ads (and as a result, lower CPC and CPA, and higher CTR)
The ability to optimize your account systematically (because it's actually organized)

You, right now

The best campaign structure: Alpha-Beta

There are two optimizations that generate the best results.

They are SKAGs and the Alpha-Beta campaign structure.

I didn't invent them - that was Brad Geddes, the longest-standing SEM professional and the best PPC resource I know of. They were later popularized by Q3 digital, one of the best PPC agencies in San Francisco.

The Alpha-Beta campaign structure splits your campaigns into two categories:

Maximum Performance (Alpha)
Data Collection (Beta)

This means every Alpha campaign will be optimized to the max, generating clicks, leads and sales at minimal cost and with positive ROI.

And the Beta campaigns will be “exploration” campaigns, where you'll discover new winning keywords for your Alpha campaigns.

This structure is basically Keyword Mining: panning for gold (Alpha Keywords) in the mud (Beta Keywords).

To implement this, the first requirement is that you're tracking conversions on your website.

Warning:

DO NOT SPEND MONEY ON SEM UNTIL YOU'RE TRACKING CONVERSIONS

For the love of God.

Without conversion tracking, it's impossible to optimize your funnel. If you need help setting this up, talk to me.

Assuming you already have Adwords and Analytics properly configured, there are two possibilities:

You're starting your Adwords account right now and have never advertised before
You already have an Adwords account up and running

The only implementation difference between these two is that if your Adwords account is already in operation, you probably already have the data to build Alpha campaigns right now.

If it's brand new, you'll need some time to test the Beta Keywords before implementing the rest.

If you enjoyed this post, you're probably interested in the topics I write (and read) about.

I share a new post every Monday, along with interesting articles and books.

Give it a try and see if you like it:

How to build the Beta Campaigns

The secret here is the keyword match types

Since Beta keywords are exploratory, we don't want them to be exact, but we also don't want them to be too broad;

You will never use Broad Match in a performance-driven account.

Because broad match wastes your money on useless search terms that have nothing to do with your product.

Google's definition of broad match goes like this:

A keyword setting that allows your ad to show for that keyword or variations of it. An ad set up for the keyword “low carb diet” in broad match will show if someone searches “carb-free foods”, “low calorie recipes”, “mediterranean diet plans”, “low carb diet program”.

You should use Modified Broad match to explore new keywords. That way, you get a bit more control over what you're advertising on.

If you advertise, in modified broad match, on “+low +carb +diet”, it will only show for searches that include all three words - so, from the examples above, only “low carb diet program”.

Let's say you teach Yoga online and did some keyword research. You picked 6 keywords for your tests:

learn yoga online
learn vinyasa yoga
yoga classes on the internet
yoga classes for beginners
best vinyasa yoga teacher
online yoga teacher

Then you'll build your beta campaign with those keywords in Modified Broad match

First, we create a campaign with a standardized name. Don't forget, the name matters a lot. It will help you automate and organize your account once it gets big.

A good name for the campaign would be:

Search - Brazil - Beta - Yoga

First, the campaign type (could be Display, Search, etc..).

Second, the geographic location.

Then, the campaign category (Alpha or Beta).

Last, the shared keyword (yoga) is right in the name, making the campaigns easy to recognize.

After that, we create three ad groups, splitting them by shared semantics:

Ad Group	Keywords
learn vinyasa yoga online	[learn vinyasa yoga online]
how to learn yoga online	[how to learn yoga online]

Enable the campaign and…

Wait until you have enough data.

Depending on your budget, it can take a few days.

Now, we start optimizing!

Pause and Negative the bad terms

Go to the Search Terms Report and sort the terms by number of impressions.

You will most definitely find Search Terms that make no sense for you. Zero purchase intent. You'll find some with tons of impressions and no clicks at all.

You should put these terms in a negative keyword list, so they get eliminated from all your campaigns.

For specific instructions, check out the step-by-step guide on how to implement the Alpha-Beta structure (coming soon)

Put the good keywords into SKAGs in the Alpha Campaigns

In that same Search Terms Report, you should also find good search terms. Above-average CTR, a nice number of conversions, and so on.

These search terms are our gold.

You should now put them into an Alpha campaign, inside SKAGs.

What are SKAGS?

SKAG - Single Keyword Ad Group - um grupo de anúncio com somente uma palavra chave.

One of the most Granular Campaign structures out there, SKAGs still spark plenty of debate in the SEM world. They guarantee Ads with sky-high quality and a lower cost per result, but they can be labor-intensive and hard to maintain.

For small and medium accounts that prioritize ad performance and ROI, I believe SKAGs are the best option.

How to implement the Alpha Campaigns:

First, we create the Alpha campaigns with a standardized name:

Search - Brazil - Alpha - Learn Yoga

Search - Brazil - Alpha - Yoga Teacher

Search - Brazil - Alpha - Yoga Classes

Because of the granular ad groups, it's recommended to split the Alpha campaigns.

And inside the campaign, we'll create ad groups with the good keywords from our Search Terms Report.

Let's say the winning search terms you extracted from the keyword “+learn +yoga +online” were: “learn vinyasa yoga online” and “how to learn yoga online”.

Now, the ad groups will have exact match keywords:

Ad GroupKeywordlearn vinyasa yoga online[learn vinyasa yoga online]how to learn yoga online[how to learn yoga online]

Now, since your ad group has only one keyword, you can write a very specific Ad - this will boost your CTR, and your ad Quality Score lands between 9 and 10.

For keywords that perform extremely well, you should even consider building a dedicated Landing Page.

Negative the Alpha Campaign keywords in the Beta Campaigns

You want that when someone searches “learn vinyasa yoga online” on Google, Google shows your specific ad, with a high Quality Score.

But that keyword is still triggering the Beta campaign we set up - so Google might show the generic ad with a ~meh Quality Score.

To keep that from happening, put all the exact match keywords into a negative keyword list - Alpha Keywords.

Exclude the Alpha Keywords from your Beta Campaigns.

You've just finished the best campaign structure Adwords can offer.

The process of generating more Alpha keywords and negativing terms - known as account hygiene - should be repeated once a week or every 15 days.

By following this system, you should be able to keep a profitable Adwords account without excessive maintenance.

Assisted Conversions in Analytics: find out what they are and use them like a Pro

Wed, 17 Oct 2018 00:57:00 GMT

Find out why the Google Analytics Assisted Conversions Report is one of the most important ones and learn to use it like a pro.

If you have a website that's recording conversions of any kind through Google Analytics, you HAVE to know the Assisted Conversions Report.

It's how you find out which type of content matters most in your funnel. And which pages actually make a difference ($$) at the end of the day.

What are Assisted Conversions?

An assisted conversion is counted when some dimension (pages, acquisition channels, ads, etc.) measured by Analytics is part of a user's conversion path, but not their final interaction.

Sounds complex?

If you enjoyed this post, you'll probably be interested in the topics I write (and read) about.

Every Monday I share a new post, along with interesting articles and books.

Give it a try and see if you like it:

Let's simplify:

Suppose you're interested in traveling and visiting Spain. 🇪🇸

You search Google for: “Best cities to visit in Spain”

The first result is from www.agenciadeviagensexemplo.com.br, and you click on it.

You read the article, and decide you have to see Madrid, Barcelona, Valencia and Granada.

Then you close that tab and go do something else.

A week later, you decide to buy the travel package that www.agenciadeviagensexemplo.com.br offers, so you search for “agencia de viagens exemplo comprar pacote”, click the first result, which is an Adwords ad, and make the purchase.

Assuming our example site has Analytics properly set up (talk to me if you need help doing that), we'll have recorded a conversion along with the purchase.

For Google Analytics, the conversion path you took was the following:

(1st) Google Organic >> (2nd) Google Adwords (Conversion)

So, we have:

an Assisted Conversion for the Google Organic channel
a Last Click Conversion for the Google Adwords channel

Back to the concept: an Assisted Conversion is any point of interaction with a Google Analytics dimension that was not the final interaction before a conversion.

How to use the Assisted Conversions Report like a Pro

You can use this report to find out which organic content is essential to your sales funnel, for example.

To discover where people find your site for the first time.

Or even to run an analysis of how important mobile is to your conversions.

Shall we?

First, to access it, go to Conversions > Multi-Channel Funnels > Assisted Conversions

*PS: My Analytics is in English. I'll always include the Portuguese equivalent of each image.

We land on the following page:

Here are the most important elements of the Report:

The Conversion Type

By default, all conversions come selected. So pay attention: if you only want to see a Lead conversion, or a sale, make sure it's the only one selected.

The Lookback Window

This is the window of past time you want to include in the report. You should adjust it according to your customer's buying journey. Usually, for B2B products, 90 days is the best option.

The Analysis Type

This is one of the most important settings.

You can choose the interaction type: Assisted, or First Click.

First click interactions are especially useful for learning more about the top of your funnel.

They tell you how people hear about you, and where your converters come from.

First click interactions simply attribute the full value of a conversion to the first touchpoint the user had with your site.

Choose assisted / first click conversions according to your goal.

The Primary Dimension

This is where the real power of this tool hides.

The default view is by Acquisition Channels. It can be pretty useful, giving you an overview of which channels are important to you.

But it doesn't give you enough detail to understand what's really going on.

So you can change the Primary Dimension to dig deeper.

For example:

Let's find out which Landing Pages are driving the most assisted conversions:

And…

We get a list of the pages bringing us the most assisted conversions.

Now, use your imagination and test other dimensions for more insights about your site.

Add a WhatsApp Button to Your Site in 2 Minutes (Free HTML Generator!)

Thu, 11 Oct 2018 20:49:00 GMT

Link and Button Generator for WhatsApp. Works on WhatsApp Web and Mobile. HTML Code.

1 WhatsApp Number 2 Widget Header 3 Message 4Code:

Check out my new projects: 1) automatically screen resumes with https://magicpeople.ai/ and 2) to track prices on your favorite site: https://price-tracker.io

How does it work?

It's just some simple CSS using media queries to show two different links: one that opens WhatsApp Web on bigger screens, and one that uses the WhatsApp API on phones.

Wordpress:

I'm going to make a plugin that adds this automatically very soon (I'd say in a week or two.)

For now, you can paste the code into your site's footer:

Appearance > Editor

Go to the footer.php file and paste it at the very top.

Don't change anything else in the file. Don't delete or replace anything unless you know what you're doing.

Don't forget to hit update on the file.

The Art of Profitability - Book Summary and the 23 profit models

Tue, 04 Sep 2018 13:05:00 GMT

The book describes 23 profitable business models that the protagonist, Zhao, identified. It teaches through parables, telling stories of companies that applied each model to generate profit. Easy to read, but it's dense, full of complex ideas that need to be revisited several times to truly absorb them.

Book: The Art of Profitability

Author: Adrian Slywotzky

Rating: ¹⁰⁄₁₀

Date Read: May-2018 - September-2018

Description:

Fantastic. The book is so good I read the whole thing in a single day. It describes 23 profitable business models that the protagonist, Zhao, identified. The book teaches through parables, telling stories of companies that applied each model to generate profit. Easy to read, but it's dense, full of complex ideas that need to be revisited several times to truly absorb them.

I liked the models so much that I'm going to take detailed notes on each one of them, along with the homework Zhao assigns.

PS: This article will be updated as I take the notes for each model

Customer Solution Profit: Costumer Solution Profit

Lose money for a while, as you find and develop the specific solution for a customer. Once your service is integrated into their operation, profit.

To illustrate this profit model, Zhao tells us the story of a company that sold financial information to banks and was having profitability problems, Factsect.

This company generated $20 million in revenue with just 30 employees. Its competitors generated $40 million with 400 employees.

Factsect was an extremely profitable company.

They operated like this:

First of all, they sent a team of 2 or 3 people to work with their clients. That team spent as much time as needed learning everything they could about them. How their operation worked, which systems worked and which didn't, and what the real need the company faced was.

With that information, Factsect developed a tailor-made solution for that client. They invested a lot of time and money building a product that would solve all of the company's needs and could be integrated into their work environment.

They invested a lot of money in this phase, and billed almost nothing.

Until the moment the product was ready and started being integrated into their client's day-to-day. The solutions were so well thought out and specific that they generated massive value for the client, increasing the company's revenue.

And at the same time, the three people who had spent all that time figuring out how to integrate it were no longer needed - with just someone handling maintenance, Factsect kept its clients happy with a single employee.

Companies that use this model masterfully: Palantir(?). Send me an @ on Twitter if you can think of more.

If you enjoyed this post, you'll probably be interested in the topics I write (and read) about.

Every Monday I share a new post, along with interesting articles and books.

Give it a try and see if you like it:

Pyramid Profit: Pyramid Profit

No, this has nothing to do with pyramid schemes like TelexFree and BitConnect.

It's a very interesting profit model, and easiest to spot in retail.

To illustrate it, Zhao uses Mattel.

Mattel sells all kinds of toys, one of the most popular being Barbie dolls.

A standard Barbie doll goes for around $30.

So, we have a product line at a mid-range price:

But if Mattel isn't careful, its competitors can produce cheaper dolls and start winning market share.

So you build a wall to stop your competitors from doing that.

Mattel develops a $10 Barbie doll. It's barely profitable, but it keeps other companies from stealing its slice of the market.

Now, we have one product line at a mid-range price, and another at a low price:

Now that Mattel has established mid-priced and low-priced product lines.

What does it go after?

Luxury products.

Barbies with much higher margins, at $300.

In the end, you wind up with a carefully planned system to defend yourself from competitors and generate profit at every price point.

Companies that use this model masterfully: Apple, car manufacturers.

Multi-Component Profit: Multi-Component Profit

A pretty common profit model. It should work for most B2C businesses.

As an example, Zhao tells two stories, one of them about Coca Cola.

The secret of this business model is having many business components. With just one product, you can have several businesses. Some of them very profitable. Others, not so much - they may serve another purpose.

Coke, for example, has a few components:

Sales in restaurants
Sales in supermarkets
Sales in vending machines

Each of these components has a different profitability.

Think about how much you pay for 100ml of Coke in each situation.

At a restaurant, a can goes for around R$5. That means 100ml costs R$1,51

At a supermarket, you buy 2 liters for 8 reais. That means 100ml costs R$0,40.

In a vending machine, it's around R$3,00 - R$0,90 per 100ml.

The difference between multi-component profit and pyramid profit is that the consumer buys the same product at every price point - the same product, several different businesses.

Switchboard Profit: Switchboard Profit

One of the most elegant models in the book, and personally, one of my favorites.

The E-Myth

Thu, 31 May 2018 13:37:00 GMT

The book describes the process the average entrepreneur goes through when starting their business, and emphasizes systems thinking to make your company scalable.

Notes taken:

Every person has distinct personalities operating inside them
a good example of this is the Fat Guy vs Skinny Guy. (One day, you notice you're fat and start changing your life. A mental paradigm shift is happening, where the Fat Guy role exits your mind and the Skinny Guy role takes over. These two personalities hate each other, and they're fighting for control. One of them will win. Usually, what happens is that the Skinny Guy takes the reins for two days, and then the Fat Guy comes back to wreck everything.
In the world of entrepreneurs, we have three roles: The Entrepreneur, The Manager, and The Technician.
The Entrepreneur lives in the future, and plans things with vision;
The Manager lives in the past, and organizes everything they see; they're averse to change;
The Technician lives in the present, and executes, does, executes.
The big problem with small businesses — which is usually a person who is 10% entrepreneur, 20% manager, and 70% technician — is that in a moment when the entrepreneur takes command, they decide to open a business. But soon the technician wins the mental battle again.
So small businesses run with almost no entrepreneurial vision, no management, and lots of action. (Also known as the //dumb// way of doing things)
Companies have three stages; infancy, adolescence, and maturity;
Most companies will never make it past infancy, and most of the ones that get past infancy will never reach maturity (90% of companies close within 5 years; of the 10% that survive, 80% close within the next 5)
Infancy is when the whole business is centered on you. *** ??? ***
Adolescence is when you bring in a manager and start expanding. *** ??? ***
I'll revisit the part about the Infancy, Adolescence, and Maturity of a business. A table is the best way to understand that information.

The idea of the three roles inside an entrepreneur's mind keeps intriguing me more and more. Picturing yourself as a fragmentation of several personalities (like in the movie Split, minus the psychopathy) seems like an interesting mental model to me.

Usually, people see their ego as something singular and indivisible - I am this person, I like bananas, apples, I'm good at this and that.

But the truth is we all put on and take off masks and costumes all the time. I'm not going to behave at a bar the way I behave at work. The person writing here is not the same person who chats with my grandma.

I mean - we naturally take on all sorts of roles already. What starts to happen if you create these characters deliberately, and choose “today I'm going to be Gabriel the Programmer”, or “Today I'm going to be Carnival Gabriel” at the start of the day?

At the very least, it would make me think about which routine I'm stepping into that day, and it might even prevent bad habits;

It's a thought I still have to develop, and I have a really good feeling about it.

Anyway, here's the rundown of what I read yesterday:

Extraordinary businesses don't expect their results from people - they expect their results from processes, from Systems.
Think of your business as a product, as if you were going to franchise it, so that it can be operated in Vladivostok just as well as in Piracicaba, the same way, run by wildly different people, who achieve the same result (McDonald's)
Your first store, or unit, (whatever it is), should operate as a franchise prototype; that's exactly what you want to build, a franchise prototype, which should be the basis of operation for the next ones (even if you're not going to work on a franchise model)
Every business you build should be an aggregate of systems that generate consistent results and experiences.
Build systems that generate extraordinary results with people

The book keeps unfolding, and makes lots of interesting points about metrics, goals, etc. Apparently there's a 12-step process for building a franchise prototype.

The first step is identifying your personal goals. Then, he starts talking about business specific goals, which threw me into an Excel frenzy, projecting revenue volume and profits for that services company. It was really interesting to use Sivers' thinking model - a business plan is a conjecture; do it now with no money. Now with lots of money. I built two models - one thinking of a one-time service, and another thinking “What if it was a SaaS?” I believe in the end the two models will coexist, with demand both for regular cleaning and for last-minute cleaning.

The process of putting together the company's org chart as the very first thing you do - starting from the end - struck me as wonderful. Of course this would never work for a disruptive startup, but I can clearly see it working for simpler businesses - and, why not, for Estoicos.com.br and the services company I intend to set up?

It works like this:

Start from the end. Picture your company's structure once it's at the size you're planning for.
Make an org chart with all the roles in the company:
That structure will exist from day one. The only difference is that you fill every role at first.
Write a description of the responsibilities and activities of each role, as well as the standards against which that person's work will be evaluated.
At the end, make a contract, to be signed both by the company's CEO and by the subordinates - if you're all of them, sign everything. Making a contract with yourself, from yourself.

The Operating System

Sat, 28 Apr 2018 17:03:00 GMT

If you make decisions without really knowing why, and regret them more than you should, you'll probably be interested in what I'm writing here.

Or if you catch yourself doing things you never even decided to do.

Before you know it, you're living life on autopilot, leaving aside the things that really matter to you.

Yeah. I have these problems.

And for a while now I've been thinking about how to deal with it…

More specifically, thinking about thinking itself - about the way we face the world, and how that changes the course of our lives and our day-to-day.

And I came to the conclusion that it all originates in my operating system.

The Operating System

Imagine you just bought a new phone.

But this phone came without an operating system installed.

The phone would be useless if you couldn't run Android, iOS or even Windows Phone on it.

The operating system is what will let you:

Interact with the hardware (camera, touchscreen, etc.)
Download apps to solve specific problems.

Now consider that you are also made of hardware and software.

Your hardware is your body: your brain, your lungs, your hands, etc.

Your operating system?

It's your mind - conscious and subconscious.

Let's assume most people come into the world and spend the first years of their lives building an operating system that lets them interact with their own bodies.

So in childhood you learn to crawl, then to use your legs, and soon you're running around, using your whole body in complex ways.

One of the first abstract concepts you learned was language.

Language is one of the first “apps” you installed on your operating system so that you could communicate with other human beings.

And soon you went off to school, which is childhood's “app store” - where you learned to read, basic math operations, etc. etc., and most likely later found yourself in college installing much more complex “apps”.

If you enjoyed this post, you'll probably be interested in the topics I write (and read) about.

Every Monday I share a new post, along with interesting articles and books.

Give it a try and see if you like it:

Now: unlike a phone running Android or iOS, where you know exactly which apps are installed, can access them easily, and move fluidly between all sorts of functions -

OUR MIND IS A MESS!

It's more like that Windows XP PC your aunt uses, with thousands of viruses and an Internet Explorer with more toolbars than browsing space, than anything organized and efficient.

And that can't be good for you…

Let's now call this operating system of the human brain by the name it should be called:

Philosophy

Philosophy is the way you think, it's how you face the world, and how you choose to employ your skills.

Philosophy is usually associated with unintelligible, impractical philosophical treatises.

Anyone who knows me knows I like to set the extreme opposite of unintelligible and impractical as an ideal for my life.

Well then: in truth, it's much harder to do anything without philosophy. It's necessary in order to be lucid and effective.

We already know there's no escaping our operating system. It's necessarily there, controlling how we interact with the world.

Now, the choice you have to make is this: do you want to live with an inefficient, confusing operating system full of “viruses”, or do you want to live with an operating system that works well?

As a human being, you have no choice about the fact that you need a philosophy. Your only choice is whether you define your philosophy through a conscious, rational, disciplined process of thought, through a scrupulously deliberate logical system - or whether you let your subconscious accumulate a junk heap of dubious conclusions, false generalizations, undefined contradictions, undigested slogans, unidentified wishes, doubts and fears, thrown together by chance, but integrated by your subconscious into a kind of mongrel philosophy - all fused into a single, solid weight: doubt in your own ability, like chains binding you where the wings of your mind should have been.

As a human being, you have no choice about the fact that you need a philosophy. Your only choice is whether you define your philosophy by a conscious, rational, disciplined process of thought and scrupulously logical deliberation—or let your subconscious accumulate a junk heap of unwarranted conclusions, false generalizations, undefined contradictions, undigested slogans, unidentified wishes, doubts and fears, thrown together by chance, but integrated by your subconscious into a kind of mongrel philosophy and fused into a single, solid weight: self-doubt, like a ball and chain in the place where your mind’s wings should have grown.

Ayn Rand, Philosophy: who needs it?

Living a life of unconsidered, random decisions will probably result in unhappiness, unfulfillment, disillusionment, and feed toxic feelings whose origin you can't even identify.

If at this point I still haven't managed to convince you that defining a philosophy for your life is really important, and that it improves your life significantly, I give up.

If you enjoyed this post, you'll probably be interested in the topics I write (and read) about.

Every Monday I share a new post, along with interesting articles and books.

Give it a try and see if you like it:

How to install an operating system that works.

This isn't very easy, but it's our only option.

The most successful people in the world have already done it.

Basically, the best way to install a good operating system is by copying those people.

Read what they read. Read what they wrote. Listen to interviews with them or about them. Read biographies.

That way, you'll understand the way they think (their operating system). And then you can decide for yourself how to apply that operating system.

You can also take a school of thought and adopt it as an operating system. (Just be careful to understand how to think, not just copy a thought without understanding its premises)

To this day, the best operating system I've found for my productivity and well-being is Stoicism.

I'm working every day to build it as best I can, but it's still pretty flawed.

I've found a few things that work reasonably well for me in some areas, but in others I'm still learning and iterating to figure out how best to behave.

Stoicism as an operating system

External things

Amusing ourselves to Death

Fri, 06 Apr 2018 21:23:00 GMT

Yesterday, I finished reading “Amusing ourselves to Death”, by Neil Postman.

It really is a great book, more current than ever. I'd like to do an analysis of it versus the modern world, which although in many places has gone back to being typographic (twitter, email, whatsapp, etc.), still seems invaded by the imagery of televised media. And even better, Netflix: what that means for how we construct our reasoning.

Huxley and Orwell have opposite visions of what a dystopia would be
Orwell's image represents a totalitarian dystopia, which happened in the soviet countries; the state takes control of information in such a way that people don't know what is true and what is not;
Huxley's image represents a 'capitalist' dystopia (closer to a welfare state), in which people are controlled by the excess of information, and by the banality of the human species
Neil Postman argues that the American society of the 80s is the realization of Huxley's nightmare.
One of Postman's most interesting arguments is the transition from a society whose main means of communication is the written media, to a society that has televised media as such;
Postman says these two types of media have two types of language - Written: ordered, leading to an end, logical. Televised: non-linear (it's made so that anyone can understand what's going on in 20 seconds), aimed at entertaining, with no necessary logic behind it: emotional
A televised language is harmful to a democracy, because then politicians are chosen for their ability to stir emotions and for their image, but not for their logic and good arguments.

How to make a city and state field in Gravity Forms

Wed, 07 Feb 2018 00:49:00 GMT

Step by step guide to adding a list of Brazilian cities and states to Gravity Forms

Over the last few months, I've been working on increasing traffic and lead generation for Delta Containers.

We're using Gravity Forms, a very popular forms plugin for wordpress, to capture leads.

One of the problems with Gravity Forms is that it has no native support for location fields in Brazil.

To solve this problem, you can make use of the very handy script cidades-estados-js (see the repository here).

Here's how it turned out:

Step by step:

First: I'm assuming you have Gravity Forms installed and working. If you don't, see how to do it here.

Then, create a form with the data you want to capture. In my case, I added First Name, Last Name, E-mail, Phone, State and City

The State and City fields must be created as Standard Fields - Drop Down.

Add a single option to both, with the same name as the field (city and state)

Save the form and click Preview

Right-click and select the "Inspect Element" option. Use the selector to find the field's ID. Save the IDs of the State and City fields. In my case they were input_13_15 and input_13_16.

Close the preview page and go back to editing the form.

Now, add a Standard Field - HTML. It should be the last element of the form.

Paste the following code, replacing 'ID-DO-CAMPO-DE-ESTADO' with the ID you saved earlier.

<script src="https://www.adamante.com.br/validator.js"></script>
    <script>
            new dgCidadesEstados({
                estado: document.getElementById('ID-DO-CAMPO-DE-ESTADO'),
                cidade: document.getElementById('ID-DO-CAMPO-DE-CIDADE')
                });
   	</script>

PS: I'm hosting the script on this page. You're welcome to use it. If you prefer, download it and host it on your own server.

Update the form and test it with Preview again.

It should be working! :)

The Eternal Tupiniquim Regress

Tue, 10 Mar 2015 20:59:00 GMT

The philosopher Friedrich Nietzsche formulated the concept of the Eternal Return.

According to him, the Eternal Return has to do with the possibility of our existences repeating themselves infinitely, forcing us to live our lives exactly the way we are living them now, making the same mistakes and the same right calls, and adding nothing new to them: we would be, therefore, trapped in a prison of repetition. Far from wanting to discuss philosophy or metaphysics here, if we apply this idea to understanding our politics, we can coin a term for what is almost like a (tragicomic) disease afflicting Brazil's statesmen: the Eterno Retorno Tupiniquim — the Eternal Tupiniquim Return.

It's almost as if we were doomed to always repeat the same mistakes of the past: we, as a State, always fall into the same traps, always resort to the same populisms (dressed up in different colors), always reach for the same disastrous economic tricks, and fail, both as individuals and as a society, to stop them from happening.

So the crisis we're facing now, personified in the mistakes of the (also tragicomic) figure of Dilma Rousseff and her party, is the same crisis we faced in the 80s, the one we faced with Getúlio, the one we faced with Rui Barbosa: it's the crisis of reckless statism, the crisis of political megalomania, the crisis of the lack of freedom. How, then, do we break out of this vicious cycle and leave behind the Eternal Tupiniquim Return — which we can't quite tell is more of a sad samba or a disastrous tango?

The answer isn't that simple. A liberal, orthodox shock to the economy, which looks like the solution, has never been fully carried out in our country. It was always done halfway, never truly liberal. To fight the Encilhamento Crisis, generated essentially by the indiscriminate issuing of money and credit, Campos Sales fought inflation by pulling currency out of circulation and blocking its issuance, pursuing a strong, stable currency. He also managed to cut government spending and shrink the budget deficit. But he raised taxes significantly, generating popular discontent. In the end, he balanced the budget, and Brazil even got room to breathe. But not for long: it wouldn't take much time for the state to spin out of control again, going through insane interventionist ventures like the Taubaté Agreement and the Coffee Valorization plans, one of the causes of the deep recession we had in 1929. The problem has its roots in the freedom politicians had to concoct their far-fetched schemes and to be reckless in running the public administration.

The story of the crisis we're living through now isn't all that different. After seven disastrous economic plans to contain inflation and generate growth, all of them with effects opposite to what they intended, a left-wing government finally gave in and implemented the so-called “neoliberal” economic reforms outlined in the Plano Real. The problem with Brazilian “neoliberalism” is that it, too, was implemented halfway. We privatized state-owned companies (sloppily), but kept their sectors heavily regulated and shielded from competition. We stuck with a model of extortionate and inefficient taxes. We spearheaded a bureaucracy so maddening that, next to it, Kafka's bureaucratic nightmares look like bedtime stories. The great triumph of the Plano Real — a relatively stable currency and relatively prudent monetary policy (in the view of economist Leandro Roque) — was soon reversed by the use of public banks to expand credit and, later, made official under the Dilma administration through the Nova Matriz Econômica, basically a Keynesian paradise. And once again we found ourselves stuck in the mud. The problem, after all these years, is still how easy it was for politicians to be reckless in running the public administration, chasing their hallucinations of grandeur.

The question is: if we implement liberal measures at this point, we would probably still be subject to the Eternal Regress of Statism. The only way to keep that from happening is by institutionalizing these reforms, preventing them from being undone, enshrining them in the constitution; it's outlawing recklessness. We should look at how this was done in New Zealand and try to adapt some of those reforms to our country: New Zealand, in the 1980s, found itself with no great economic prospects, stagnant, with a very closed, heavily state-run economy. Nearly irreversible reforms were made, mirroring the thinking of the Chicago School and of Friedrich Hayek, who sees the market as a discovery process and stresses the importance of information flows for the economy to become more rational. Specifically, you can see this implemented in the public sector, which became far more transparent and predictable, not only in its actions but in its targets and results, making planning based on those variables much simpler for the productive sector. Laws were also created forcing politicians and bureaucrats to be more careful and transparent with public finances, like the Public Finance Act, the Fiscal Responsibility Act, and the Central Bank Act (this last one deserves a dedicated article to explain it, given how unique it is) — laws that can't be changed on a whim, the way we do here with, for example, our own Fiscal Responsibility Law. Today, the country is one of the most prosperous and free economies in the world.

It's obvious we have countless reforms to carry out, and that they still need plenty of discussion, but we should take advantage of a critical situation like the one we find ourselves in now to implement measures that would hardly be feasible (for political reasons) in calmer times. We also know that liberalism in Brazil is demonized by the left, but liberal thought has been growing consistently and encouragingly, and lately several politicians and a few parties have emerged supporting ideas of freedom: after all, we don't need to change Congress to get support for these reforms — we need to make our ideas politically profitable, as Milton Friedman put it. And in fact, we're pulling it off. And maybe we're getting close to the end of this prison of political megalomania and irresponsibility that is the Eternal Tupiniquim Regress.

Gabriel Adamante

Token Split Tool - Easily Split text based on token count

Output:

NFe Analytics: analyze data from NFe XML files, free and easy

Finding the ideal apartment (Like a Hacker)

#1 - Defining the specs

#2 - Searching the obvious places

Extracting the data from the platforms

#3 - Striking Gold

Finding every local real estate agency

Wrapping up the saga

Histórias de Fim do Mundo (End of the World Stories)

The 6+1 Copywriting Formula: (sell, don't push)

Step 1: Context

Step 2: Attention

Step 3: Desire

Step 4: The difference

Step 5: The Solution

Step 6: Call to Action

+ 1 step: Credibility

Book Summary: The Power of Thinking Outside the Box: The Lessons of 8 Executives Who Overcame Seemingly Unsolvable Problems

The book is a case study of 8 CEOs of public companies who achieved extraordinary returns in their companies' stock value, and isolates the factors that led them to such success.

How to run a Product/Market Fit survey (the Sean Ellis Survey)

How to implement this survey:

How to Learn Growth Marketing

What is Growth Marketing? Or Hacking?

How to learn Growth Marketing?

Growth Team vs Growth Professional

Learning about user acquisition and conversion

User Acquisition

User Conversion:

Be a unicorn: pick your Marketing Stack

Testing like a pro: continuous experimentation

What is an A/B test?

What is a channel test?

The Guide to Attribution Models in Google Analytics

Which click was responsible for the sale?

Last Click Attribution Model

First Click Attribution Model

Coming soon: Linear, Last Non-Direct Click, and Position-Based Attribution Models

How to use the Google Analytics Attribution Model Tool

1) Conversion Type

2) Analysis Period

3) Attribution Model Selector

4) Secondary Dimension

5) Date Selector

Remote Work: the good, the bad, and the ugly

Productivity

The Good

The Bad

Freedom

The Good

The Bad

Social

The Good

The Bad

Whatsapp Button Generator: put on your website in 2 minutes - WhatsWidget

Put a WhatsApp button on your website in 2 minutes: generate the button and paste the HTML code

How to use:

How does it work?

Wordpress:

Aggregators and Platforms will swallow the world

What do Apple, Amazon, Alphabet (Google), Facebook, Microsoft, and Alibaba have in common? They are aggregators and platforms.

The Complete Guide to Conversions in Google Analytics

A guide that goes from installing conversion tags using the main methods, all the way to complex topics like conversion attribution models.

Contents

What are Conversions?

Why Conversions matter

How to install Google Analytics Conversion Tracking

How to pick the best tracking method:

Conversions via PageView:

Step by step to Set Up a Pageview Conversion in Google Analytics

Conversions via Events:

Setting Up Event Goals in Google Analytics

Sending Events with Google Tag Manager

Sending Events with WordPress Plugins

Sending Events with JavaScript

Importing Sales data into Google Analytics

Recording a user's ID in Analytics:

Sending the conversion data to Analytics through the Measurement Protocol