As we officially set into the New Year, I’ve been thinking about how SEO is changing and where we are headed.
One theme I’ve been hearing over and over again is automation.
So how can we make automation actionable? What should we be doing to actually automate processes?
The answer is Python.
Today’s guest is a pioneer of Python, Hamlet Batista, founder of RankSense. He has invested a lot of time and research into Python and devotes numerous talks and articles to the subject.
In this episode, we dive into what Python is, how it can help agencies as well as the ominous learning curve and the best beginner resources.
You have a slightly different background than most of our guests. Can you tell us a little bit about yourself?
He is a trained engineer and learned to write code, but his first business was in marketing. He started as an affiliate marketer and accomplished high rankings on Google. His engineering background helps him to fix inefficiencies.
What do you do with RankSense? You have a number of patented SEO tools.
RankSense helps you implement your recommendations. When you think of marketers, they typically create recommendations and then have to carry them out. On the technical side of SEO, you don’t have the flexibility to go in and fix them yourself. You have to send it to a developer.
However, that limits what you can do (ie redirects).
That’s where RankSense comes in. It can do everything tag manager can do in addition to redirects etc.
You believe that marketers should also have some development skills. Can you elaborate on that?
He’s been talking a lot about Python and use cases regarding Python.
Developers and marketers operate in silos.
Developers are taught to take away generalize and remove context whereas marketers understand the psychological side of why people do what they do.
This creates a gap.
There are stories hidden in these gaps that you as the marketer could uncover them, but only if you can extract the data yourself. Your developer will miss it.
Can you give an example?
There is a visualization, Data is Beautiful, on Reddit. The visualization was made by engineers/scientists and it is supposed to show how to maximize your time at Disney based on how long the rides last and how long people will have to wait.
But there’s a problem.
As a marketer, I know that wait time and ride time are not the only factors that go into my overall level of enjoyment. It’s also the quality of the ride. And the quality varies depending on if I am a woman/man, child/elder, etc.
However, engineers/scientists miss this part of the story.
As a marketer, I am going to break this down and include these other variables that I as a marketer understand. Now that it is broken down, this info is useful for say, a travel agency.
Prior to that, a travel agency would have been forced to use this data with a one size fits all lens and we know as marketers that this is not true.
Get the notes from inbound here which includes visualizations and descriptions.
Can you give us an intro to Python?
Python is a programming language.
All marketers work in spreadsheets. However, most of the work that can be done in a sheet can be done in a program like Python.
Instead of having functions in a spreadsheet, you can create your own custom functions.
For example, if I want to know the best rides by time and include other pieces of data like age, it would be difficult as you would have to pull a lot of different information. However, this can be easily done with Python.
With Python, you can automate the workflow.
There’s a huge learning curve with Python. About how much time does it take for you to understand enough that you can start using it in your workflow?
A lot of tutorials are for programmers.
Start with this guide on SEJ.
This guide takes code that already exists and then I go over it line by line to understand what it is doing and learning how to modify it.
John Mueller asked a question on Twitter. He asked, what type of content performs better in mobile vs. desktop.
No tools give that info.
So Hamlet wrote a script and pulled info from Wikipedia and put it in a spreadsheet.
Learn about that example here.
The article takes the script from Mueller and it shows you step by step what he is doing. This introduces you to Python and how you can use it.
About how much time do you think that learning Python would save an SEO agency?
Hamlet doesn’t work directly in an agency, so it is difficult to tell.
At an agency, it’s about standardizations.
For example, at Adapt Partners, they use Notebooks instead of Google Sheets.
The equivalent of Google Sheets are Notebooks in Python. After you do a process manually in Sheets, you can make it into a Notebook and repeat it time and time again automatically.
There are a few people doing really great research in Python.
Ruth Everett at Deep Crawl has some great articles on SEJ and will also be speaking at Brighton SEO about it.
Hulya Coban has also done a lot of articles on advanced data analysis with Python.
There are also a lot of job postings at big brands, for example, Finder.
You mentioned that at the bottom of most of your SEJ articles, you talk about interesting use cases of Python. Do you remember any off the top of your head?
There was one where an SEO on Matt Lacuesta’s team scraped the People Also Ask Question box and created a visualization tree.
SEO belongs to the next generation because the new generation has an open mind to being more data-driven and analytical.
This is not about you becoming a programmer. This is about you acquiring superpowers to be a better marketer.
These extra skills are how you can differentiate ourselves.
Last-minute note: check out Britney Muller’s Intro to Python