A handful of lessons I learned (and I am still learning) while trying to apply #DataScience to #SEO. Some of them are not so obvious either.

This is an updated thread with new personal considerations 🧵
Communication is hard and you will get mad a lot of times. Non-technical people have no clue what you are talking about and you have to educate them.

Easier said than done, but I think that you should stay strong and keep trying.
Data quality is all. In SEO it's way harder as you are working with estimates and you don't even know the original data distribution.

That is why I am very careful when using Machine Learning models for SEO.

Now I'm getting more used to NLP tho.
SEO is challenging, now imagine explaining bits of #DataScience and SEO together to non-technical people.

Mastering the art of communication is a must, while keeping ethics into account.

That's why I recommend learning how to use PPT (no joke) or Canva to create infographics.
Visualization is your way to go for most things. People don't understand the impact of your work?

Plot it and show them your progress visually. That's why something boring like query count can be shown as something magnificent. Image
You don't need Deep Learning to solve simple tasks. You don't even need it for most SEO tasks either, barring a few exceptions.

Basic data analysis and plotting are way more useful than you think.

This is especially true for tabular data.
There will be moments where you will use DL models, that's true. In general, you will mostly rely on basic data manipulation and plotting, plus some NLP models.

If you have to learn something focus on what creates the most impact.
Mentally prepare yourself for hearing something like "OK, now can you do it in Excel?" when you are working in Python/R or whatever programming language.

After some months of work, I am finally getting less and less of this.
You have to be very careful when you define KPIs and metrics. A blog will have a different business model compared to e-commerce, implying different success metrics as well.

Even different sections of a website can have different goals/intents.
Google APIs are your best friend. Imho GSC API alone is a goldmine of information and if you have a medium-sized website you already have enough material to create crazy stuff.

Create some basic workflows and you are fine for most of your tasks.
It takes much more time to combine 2 disciplines. Consistency is key and the results will be much higher if you think about the long term.

I don't suggest this path to anyone, it's just for those who have an interest in data.
The king of data is SQL. This one is mandatory and should be improved over time.

I am trying to learn BigQuery as of now and guess what? It's a Google product

You don't often hear of SQL in the SEO community but you should learn it.
Most times you'll be frustrated because you can't use the cool and shiny scripts that you created.

It's normal, you have to think about value first, to avoid being disappointed.

Set lower expectations and try to create milestones for your projects.
Most data sources for SEO coming from tools can be considered "clean" compared to what you see in other industries.

We are so lucky to have tools with CSV exports or an API for everything. The downside is having inaccurate data.

Either way, data cleaning > all the rest
#NLP is a great friend for SEO. You have so many options to find cool insights with small scripts.

For more details:

There is a lot of hype around data, it will eventually fade away.

That is why you have to learn the basics! Most people will probably focus on buzzwords, you should do the opposite instead.

Look for momentum and strategy. >>>
>>> Let me explain, strategy as a whole is underrated. Everything is strategy. If the mass is going in one direction you should question why they are doing like that.

Don't go where there is fierce competition, differentiate yourself.
If you want to go all-in consider studying Statistics or Data Science. Beware, your focus should be on creating value first, your client/employer/whoever pays you doesn't care at all!

This is a shocking truth for people with a strong academic background.
You don't need to learn super-advanced concepts, understanding the key ideas is fine.

And I will be blunt, most concepts are so theoretical you will never use them. For SEO you can get away with knowing the essential, as long as you make good use of it.
If you're an in-house SEO try to convince your company to have a pipeline or at least a data storage. If you don't, expect to have to deal with tons of manual work and checks.

Scaling SEO processes requires a company to have some idea of how to deal with data.
I am continuing to study and summarize patents, some technical knowledge is required to be faster when skimming.

Data Science and Machine Learning are different subjects but they are strongly related. Improving your knowledge in both is highly beneficial.
Graphs are not new, even though their applications have surged in the last decade. For SEO we talk about Knowledge Graphs, one typology involved in representing knowledge.

There are other types like social networks/graphs
I am noticing that graphs are super valuable and will probably become more and more important due to their versatility.

Web 3.0 is another factor to take into account and you should already get used to this concept.
The more time I spend with data, the more I understand communication and persuasion are important.

This is something that isn't always taught in unis. Try to have in mind what your audience would like to have as an output.
You can start by replicating what you do in Excel/GSheets in Python. It's easier this way and you can see the benefits or the pitfalls of a programming language.

This is the best exercise to start your conversion to coding.
Most people think that coding is impossible and non-technical people should stick with Excel. I disagree because things will certainly change.

New generations are studying newer subjects that weren't even taught when I was in high school.
Data have impacted SEO in different ways. One example is provided by traditional tools including search intent and other suggestions based on ML.

Moreover, we have seen the rise of newer SaaS solutions based on clustering or knowledge graphs and entities.
Data analysis is one thing, automation is another. I prefer the former over the latter because we SEOs have the right domain knowledge to understand the data.

This is not true for Data Scientists without SEO experience.
Speaking of which, automation doesn't require the same amount of knowledge as analysis.

Automation is a powerful ally to save time and invest more into other activities. It's dangerous to claim that everything can be automated.

One example is given by toxic links reports.
This is a recurring topic in the SEO community and I don't think you can use tools for such a delicate process.

Sure, you can compare values across 3 or more software but what for? Disavowing is not so good as many claims to be
You can automate some steps of Keyword Research or even reporting.

You cannot automate something that requires strategy or tactics. I mean, you can try to a certain extent but you will notice that it's not always the best outcome.
Even though we are learning more and more IT skills, companies have still problems finding talented individuals that can help them.

Learning Python/R/SQL or even Power BI or Tableau will pay off!
Not all SEOs need to choose this path, I am saying this again. There are so many opportunities and markets that it would be stupid to tell you all to do the same things.

Pursue what you like and find job market gaps.
Recently, Data Engineering and cloud-related skills have increased in demand. If you want to try something different, go for it!

SEO and Data Engineering may seem distant but, as mentioned before, companies need to create pipelines or understand how to manage data.
You can take it to the next level and focus on market research for products or industries and not just for SEO.

This is very niche and just for a few selected people who have a super solid background. I have seen some professionals doing that, it's very rare.
For instance, let's say you want to launch a new app. You can do extensive Keyword Research but what if you were able to combine it with market research?

This is my personal opinion, take it with a grain of salt. Costs could outweigh the benefits.
Working with data may get super expensive and some projects are not worth it. Be sure to assess whether you have such a budget and if the problem lies elsewhere.

Imho the problem with most websites is a strategy and not even SEO.
The strategy problem will be covered in another thread. Data cannot save you if you are not able to find a good use case.

So the end goal here is understanding how to leverage what you have.
Imho, the situation can be summarized as follows:

- You get insights, you fail to act on them
- You misunderstood the problem
- You don't even have the data

... and much more! That's why I discourage working with data if you don't understand why you need them.
The strategy will always be the most important factor, followed by tactics.

What is your value proposition? Can you enhance it with the help of data?

It isn't always possible!
Next threads will focus on non-tangible skills. Please, don't ignore the fact that strategy plays a vital role in SEO.

This is the difference between having a plan and shaping a project and getting carried away. You should be proactive and not reactive.

>>>
>>> Methodology is necessary for documenting your steps and defining an effective SEO strategy (and plan).

This is a good method to fight imposter syndrome and understand where you are lacking.
After this thread you should be able to understand why data are not your top priority without proper thinking.

Methodology and strategy are your friends.

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Marco Giordano 🇺🇦

Marco Giordano 🇺🇦 Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @GiordMarco96

Mar 15
What if the problem is not #SEO related? How come, you are an SEO and you face a problem that is not connected to your job... or is it?

A thread about the importance of value proposition and quality for every SEO project 🧵
SEO is just one part of the story. It is not the only digital channel and will never be.

I prefer to focus on SEO because that's my path. This doesn't mean a business should do SEO. Being holistic is key.

That's why understanding how you want to reach your audience comes first.
Business and Marketing are not about having a good product and that's it. They cover other areas of interest and techniques that promote what you sell.

There are bad products with excellent marketing or vice versa. In some cases, they are both bad.
Read 46 tweets
Mar 14
People who are just starting in #SEO have a higher chance of meeting those pesky SEO myths and common beliefs.

This thread is for beginners and is aimed at understanding and preventing the most common pitfalls for learning SEO, plus some tips 🧵
Let's start with the basics, my recommendation will always be to start from the following free resources:

- learningseo.io
- developers.google.com/search/docs/be…

These 2 are the bare minimum and are correct. You don't need anything else to start.
The most common belief they want to sell you is that SEO is business independent. Let me explain, claiming that the subject is a set of techniques or something you can apply indiscriminately.

SEO overlaps with a lot of other subjects, it is not a watertight compartment.
Read 38 tweets
Mar 12
Using #Python for content optimization in #SEO? You must be crazy, man.

And yet, there are some cool applications I will show you in this thread 🧵
Named entity recognition (NER). Extract named entities from a text to see what your competitors or Wikipedia are using for a given topic.

This is not about keywords but the co-occurrence of specific terms.
You can do that via Google NLP API or spaCy. The first can give you a measure of the importance of the entities, called salience. The higher, the most relevant for that text.

The second one has different perks and can be trained, meaning that you can make domain-specific models.
Read 34 tweets
Mar 11
[Case study]: How I got a publisher website past 400K sessions per month with Semantic #SEO and careful planning.

This is my longest thread so far and I will try to document all the steps I followed and the main takeaways. 🧵
The niche is pop culture (actually two subsets) and the market is Italy. Zero budget as it is a test project and I am just helping a friend of mine.

Everyone is writing and the most important skill, in this case, is knowing the industry.
The first thing I did was to do a technical audit back then to spot serious issues. Since I know the niche I can tell that it's not so important unless it's dragging you down.

The technical situation of the website wasn't that bad.
Read 47 tweets
Mar 10
I've talked about Natural Processing Language (#NLP) before. What is the difference with NLG and NLU?

Behind these terms lies something more important for #SEO Specialists.

I will explain you what are these strange acronyms in this thread 🧵
For the NLP definition, check my other thread on the topic. It is a clear and concise explanation on the subject.

Natural Language Generation (NLG) can be defined as the use of Artificial Intelligence to create content.

This is what tools like Jasper.ai do. They can generate texts according to your instructions and depending on how they are trained.
Read 30 tweets
Mar 9
The #SEO world is sadly filled with misinformation. One of the many reasons is that it is a non-academic subject.

A lot of case studies lack rigid methodology and solid proof.

This thread contains my personal considerations on SEO as a whole, considering what's good and bad 🧵
Learning SEO is a nightmare. Contrasting opinions, totally different niches and markets.

The main information sources have outdated info or are just repetitive.

The best solution is right here, this is the most complete offer out there:

learningseo.io
When I was student (literally months ago) we had some meetings with ""SEO experts"".

None of the stuff presented was SEO at all. The focus was on using one tool instead of teaching you the mindset or the basics.

Learn by yourself or have a good mentor.
Read 29 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us on Twitter!

:(