Wear the data science hat more often

Begum says: “As SEOs, we must also be data scientists in disguise. We often take the data science and analysis steps unknowingly, one by one. As you progress as an SEO, you’ll advance beyond doing work for smaller websites and have to figure out how to work on bigger datasets. As the simple tasks become more complex, you might wonder how good you are at deriving conclusions.

We tend to talk about how being data-driven is incredibly important, but are we really driven by data? If we are, how good are we at analysing and interpreting it? You might be able to dip your toes into Python, mostly for data analysis and automating some daily tasks. The biggest tip for 2023 is to get ready for the SEO data revolution and wear the data scientist hat more often.”

Is there a particular process or way of working that an SEO can learn from a data scientist?

“There are five main steps to data analysis: the first being to define the questions. Let’s say you have identified a problem, or tasks to analyse your current situation and ask the right questions in order to define your goals. You can do that daily - and this might even come from your clients. Then you can collect the data to be able to present something against that and do further analysis. You can gather data from Google Search Console, Google Analytics, and other third-party tools. Taking advantage of APIs and other automation tools makes this process more efficient. Then you can clean the data and, once that’s done, you can start analysing it.

This part might derail some people from the conversation, but there are different models that you can use to derive analysis for the specific task at hand. Sometimes that takes the form of keyword research and sometimes it’ll be internal linking, etc. However, when it comes to doing this via data science, it varies from descriptive analysis to diagnostic and predictive analysis. Lazarina Stoy is a great inspiration. She has shared many common data analysis scenarios to do with SEO analysis, and offered influence for interpreting, visualizing, and sharing results. This is the fun part of our SEO processes because whether you’re using data science or not, after efficiently analysing your data you should communicate the output with stakeholders. Google Data Studio is great and can be part of it, but you’ll have to become better at it, because the data is getting better and better each day.”

How do you ask the right question to define your goals?

“This is a learning curve, especially if you don’t speak to developers or bots. It can be a difficult process but you should identify what your end goal is, get all of your data from Search Console and try to merge it with data from rank tracking tools.

You could use position tracking, for example. Combine multiple data points and identify what you intend to do within that process through data analysis. Everything will come together and you’ll find an easier and faster way to do that without using spreadsheets or any other solutions.”

In terms of collecting data, do you collect all the data available to you or do you determine specific types of data to collect?

“Getting all the data you can is one of the worst things you could do, because then you’d have to clean the data. When it comes to identifying what you want to do and how you’re going to do it, keep that goal in mind and try to find the best data sets that you can get your hands on. This is a vital step.

Let’s consider Google Search Console data. There are thousands of keywords there, and some of these you won’t really care about. Some are just typos and others are in our reports even though we don’t care about them. Excluding all of these from your data set will give you better data analysis at the end.”

What are you looking to take out when you clean data?

“It depends on the irrelevant data and the data that has errors. You should remove any data with errors or that’s repetitive. You can think of data analysis as a way you haven’t learned yet, that’s a beefed-up version of what you already do in Excel. For example, when you get your data in spreadsheets, you can just remove the duplicates and get rid of the data that’s not suitable. When it comes to data analysis, there are models you can use to get rid of error data. On top of that, you can use your analysis to determine what doesn’t serve you and get rid of it.”

What’s the best starting point for SEOs when it comes to different types of analysis?

“It depends on the analysis you’re doing because, when you’re doing keyword research vs crawling, things change. For example, you can crawl your website through Python and anything that you do there is going to be different from your last attempt or current keyword data. You can slice it down but it’s very dependent on your task.”

Conventionally, data science has been seen as more of an enterprise SEO thing. Should data science be approached by SEOs working in SMEs?

“Definitely. One would think it applies to enterprises more than it does to small and medium businesses since the data sets are larger in comparison. However, wearing a data science hat will help all kinds of businesses, more often than not. Frankly, in an SEO setting, the data science team would take

the burden of coding from you but it’s still important to deliver your story and communicate goals clearly between teams.

The best approach is to have an understanding of what can be done with the help of data science. When it comes to working with smaller businesses, you’ll mostly be doing the analysis. Data science will save you time, help you focus more on your strategy, and handle the heavy lifting so you can provide better results for your clients. Data science is something you can use regardless of the type of business you’re operating in.”

Can embracing data science help SEOs communicate with other departments in the business?

“When it comes to communication, it’ll always be very much dependent on your communication skills. However, data science can help you drive more meaningful results from your data set. That’s going to save you time that you can use to think about how to communicate things to different teams within your organisation.”

How often should this data scientist-type approach be part of regular SEO activities? Does it need to be ingrained into day-to-day activities?

“It depends on how you’re going to use it. At the moment, you might think you’re not going to use it that often - but you can use it to crawl XML sitemaps and get the static in bulk. This is better than typing everything into Screaming Frog. It can help you do these repetitive tasks much faster and more easily. If you are going to take that route and use it to automate some of your tasks, you can definitely use it daily. What you’ll want to do is analyse the data that you have at hand and then you might end up using it less regularly.”

Can this have a measurable impact on your SEO success?

“It may indirectly affect your SEO success because you’ll be able to drive accurate and meaningful conclusions from your analysis based on your data set. However, for the most part, it won’t affect your capabilities as an SEO. You can identify your goals and use data science to do what you were doing but in a faster and more meaningful way.

Human error is common and, when analysing big data sets, your mind can wander and it can become difficult to see the whole picture. Data analysis models are helping us to identify the gaps and where we are missing. When it comes to analysis and identifying the next steps, it can help you become a better SEO and more easily measure the effects of your work. Data analysis is useful for creating more time to spend on a strategy and executing or understanding data in new and better ways. You’ll also save on projects and discover new approaches for firing SEO in different ways.

What shouldn’t SEOs be doing in 2023? What’s seductive in terms of time, but ultimately counterproductive?

“Not keeping up with what’s new. Also, not linking and cooperating with other channels. It’s an evolving industry that’s influenced by recent developments in the digital sphere. That’s why it’s vital to maintain curiosity.

What we’re doing is exciting and the world continues to move faster and faster. Keep close contact with other marketing channels to help you retain an integrated strategy for your brands that will positively affect everyone involved.”

Begum Kaya is an SEO Consultant at BK Solutions and you can find her over at begumkaya.com.