5 Common AI Mistakes In The SEO Industry!

AI | Blogar

AI replaces content creators

AI is like a child, it needs text to learn.

The text comes from humans, so machines actually learn from us. Given this premise, it looks like AI can be a great writer, right?

It isn’t the case in SEO.

Many SEOs claim to rank websites with AI but never show the full consequences of doing so.

Getting less than 1K clicks per day with AI isn’t a big flex either.

It’s possible to rank AI websites but there are some considerations to make:

  • Organic Traffic isn’t forever, you will eventually lose it
  • Conversion rates are horrible
  • You get the same content as other people

Even with super valid prompts, current AI has several limits and one of them is creativity.

Now, you can tell AI to use unusual words and speak with a certain tone but the content won’t be as good as a human writer.

Even with the detailed sample above, the output will never match what I can produce manually.

Still, it’s an opportunity to get briefs, templates, and drafts!

What to do instead: Use it for briefing and to generate ideas.

ChatGPT is super innovative

OpenAI solutions were used in SEO since 2019/20 and they were content spinners 2.0.

ChatGPT isn’t the ultimate solution to building a SaaS. Some tools make API calls to OpenAI and offer 0 competitive advantage.

So you are paying for a cool interface, not for the functionalities.

This is quite common in the SEO tool industry as many tools you can clone with the right skills.

I’ve cloned a lot of them for personal use as I only needed some features.

If you are paying for a tool that simply does something and has NO unique data, then you can clone it easily.

And again, this is much before ChatGPT so you don’t even need it.

AI caused more awareness in terms of what it’s possible but this is what coding does.

Use it as an opportunity to code instead and save money.

What to do instead: Use LLM only if needed and you have no alternative. It makes 100% sense for a chatbot but 0% sense if you need to write simple code.

Content is Dead

This is part of the online fearmongering campaigns to let you purchase AI tools.

Everyone wants to use AI but how many users want to read machine-generated content?

Users first, rankings second.

By definition, if you work in content, you should be quite good at creating it.

AI is a good tool to come up with ideas and speed up some drafting but it’s not a writer!

The only content that is actually dead is the thin unoriginal affiliate content that populated the SERPs 1 year ago.

What to do instead: expand your background and analyze YouTube/Social content. AI can even help you with distribution & repurposing!

AI Can Do Data Analysis

OpenAI’s Code Interpreter is a valid tool but severely misused.

Analytics isn’t coding or creating cool plots, that’s easy.

It’s more about generating insights and finding useful patterns that translate to valid actions.

AI doesn’t tell you that, forget it.

And no, Analysts won’t be replaced because some Twitter threads claim that.

You need to know the industry and get some practice in what it actually means.

I use ChatGPT daily to write or review code and it’s quite good.

But for doing analyses? No way, it’s not as good as a trained human (or an intern, for those who ask).

Also, you should never insert personal data into ChatGPT because they will be used to train their models!

Code Interpreter is a good tool as well but ask yourself: who is promoting it? SEOs are not engineers, so you’d better check your sources before you use something 

What to do instead: Use Code Interpreter as an assistant for coding. Never feed it personal data or use it for business context.

AI moves faster

Uhm, actually not.

AI looks fast but it’s quite slow in practice.

Let’s say you have to create a tool that analyzes GSC data.

You decide to fully rely on AI and write a prompt that explicitly asks to analyze data.

I am happy to inform you that this solution isn’t optimal.

The reasons are many:

  • You still need to write code, you are only skipping a part
  • Black box approach, harder to debug and standardize
  • More costs
  • Reliance on a 3rd party (which is not good)

Developers leveraging AI move faster; making API calls and building your tool instead of writing regular code isn’t faster.

It will cause a lot of future headaches (and no value proposition).

What to do instead: Take the time to study how things actually work under the hood. It’s not your priority as an SEO but will help you in detecting issues.

©Copyright All Right Reserved 2023 | Blogar