Watch out for better, AI-powered technical SEO tools

Pam says: “Keep your eyes peeled for new, better technical SEO tools that utilise AI.

There are a lot of great technical SEO tools out there. That includes tools like Screaming Frog, that crawl and gather every bit of technical information about the site for you, or some of the more all-encompassing platforms where you can manage everything in one place, both technical and on-page. There are great tools out there now, and maybe they’re utilising AI a bit, but they’re still traditional SEO tools.

I would love to see some truly impressive AI-driven technical SEO tools. When you run a Screaming Frog crawl, you get every bit of technical data about a site you could ever possibly collect. You’re basically recreating Googlebot. Then, it’s up to your eye, and real intelligence, to figure out what to do with it.

In this age of AI, why do the tools stop there? Why can’t they use AI to go further, reverse engineer Google’s algorithms, and tell us exactly what our biggest wins would be?”

Would you like to see a Screaming Frog plugin for ChatGPT?

“Something like that would be incredible, but maybe not just for ChatGPT. Everybody is tunnel-visioned on ChatGPT – because it’s amazing and it’s breaking barriers that haven’t been broken before. However, I think it’s too limiting to constantly think of AI as just ChatGPT.

I’ve had some software developers ask me to review their new AI-driven Google Analytics tools. There are a bunch of different ones, but they’re all very ChatGPT-driven. They’re trying to mimic ChatGPT by giving you the ability to chat with your data (you can ask it how many visits you got in X time frame from Y source, etc.). It’s not that impressive to me because AI could do so much more. In the SEO and analytics world, I want tools that utilise the machine learning aspect of AI, more than the chat aspect of it.

I’d really love to see something along the lines of a Screaming Frog AI that learns from all of the SEO professionals who use Screaming Frog. It could be a SaaS software, all online, that anonymously collects all sorts of information (with consent) about how the best-performing sites are technically configured versus how the worst-performing sites are. It could understand which technical changes created the most improvement and which were useless.

Imagine all that knowledge from all those SEOs using Screaming Frog, all collected into one AI that’s constantly learning. That would basically reverse engineer the Google algorithm.”

Could that AI eventually identify where the issues are in your site, and make site performance improvements for you?

“Eventually, yes. I would not trust it to that degree yet. A lot of the AI tools that I’ve seen thus far are not that impressive. It would probably take several years of verifying with real intelligence that artificial intelligence is making the right suggestions before I would trust it to make those changes for me.

The final stage of this would be auto-tuning the site based on Google’s algorithm changes. The AI would notice that your site needs an edit because of an algorithm change, and just do it. That would be very interesting. It would become a constant, rapid, iterative race between the SEOs and Google whereas, currently, when Google makes a move and it takes us a while to catch up.”

Could you let AI loose on a relatively unpopular category in your blog and allow it to make some minor tweaks to attempt to rank it more highly?

“I would be more comfortable with an approach like, ‘Don’t fix it if it’s not broken; only tinker with the broken stuff.’ If it could be limited to that, I would feel a bit better about it. I would also want a way to roll back changes – like a really good changelog, where you can monitor it and see what the AI did and why. Then, if you don’t agree with it, you could just roll it back. I think that would be necessary.

Unfortunately, I think we’re incredibly far away from that at this point. As rapidly as AI is evolving, I can’t imagine it getting to that point in 2024. What I’m excited about for the coming year is that tools will hopefully start to come out that utilise the learning aspect of AI, as opposed to the ‘chatting with a bot’ aspect of AI.”

Are there any AI tools that you would recommend for SEOs at the moment?

“Not with regard to technical SEO tools. That’s what I’m excitedly waiting for.

As far as content goes, ChatGPT is just amazing for content ideation. However, not for writing your full content and publishing it with ChatGPT as the author.”

If AI knows exactly how to make a website perform, will all sites end up performing equally well, making content and experience more important than technical SEO?

“That’s an interesting thought experiment. I can’t think too far beyond that without thinking in circles. It’s a chicken or egg issue at that point. At that point, Web 3.0 might have taken over and we’ll all be focusing on that instead. At the moment, we have to optimize both the chicken and the egg.”

Is there anything that should never be automated?

“Content production. Content ideation – planning and strategising – is what I recommend automating with AI. The content production part should never be automated, especially considering the new E in EEAT. Google has told us specifically that they want us to add experience to expertise, authoritativeness, and trust.

I think that’s because EAT can be mimicked by AI but personal, firsthand experience can’t. ChatGPT can make it up, but it can’t create a genuine case study about something that you did for a client that worked. Nothing can replace that first-hand experience. The full content production, with the injection of personal experience, should probably never be automated.

Your process and project management, however, can be automated. I’ve implemented some things using AI that have provided tremendous improvements for us, from an automated process perspective. Tools like ChatGPT can write conference call summaries based on a transcript or create an automatic report review.

I have created a prompt in the GPT for Sheets add-on where I put all my thoughts about what I look for when I review a report before it goes out. It checks the report summary for all of those things: Did we say anything that sounds like we’re making a promise? Are we using non-committal language? Are we using relatively positive terminology, while still being accurate? Everyone should be trying to use AI for their own internal process optimization.”

Do you share with clients that you’re using AI to assist you?

“It depends on what it’s for. If it’s for their written report summary, no. It’s the same as using Grammarly to double-check emails before sending them out. For a more specific application, though, we would tell them.

For example, we had a client who purchased an old magazine that had been produced for around 40 years. The magazine had given them all the content and pictures on DVDs, and they were turning that into a website, and each picture had a really robust caption. In the process of turning this old-school print magazine into a well-optimized website, we needed image alt text tags. The captions were too long for that, but they were so accurate in their descriptions of what was shown in the image that we told them we’d like to use AI to shorten the captions into img/alt text tags.

It was tens of thousands of articles, so it didn’t make sense for a human to do it, and the cost would have been extremely prohibitive compared to using the ChatGPT API. For that application, we recommended AI to turn the captions into shorter image alt text tags for SEO and ADA purposes. We designed the project around AI’s capabilities.

You want to have that open conversation. I naturally end up having that conversation with clients when I talk about how they can use AI to come up with ideas but not to write their content. I’m constantly preaching about the responsible use of AI as it pertains to SEO. That’s my overall theory, even when it comes to client management. If you’re going to write up a report summary, and you want to use AI to slightly polish it up, that’s a responsible use of AI – as long as it was written by a human in the first place.

If you’re going to do their project work with it, they should know from the get-go. It should be used responsibly and selectively, for very specific cases where the benefits are almost impossible for a human to achieve.”

Will it be necessary to use AI to compete in the future, based on the size of projects it can accomplish?

“Yes and no. In some ways, what’s old is new again. At some point, the AI will be so overdone that businesses might seek out consultants who do things the old-school manual way with their own brains. In an AI-dominated world, the injection of real human intelligence could be the differentiating factor.

On the flip side, the scale of certain projects and the role that AI plays in scaling probably would limit competition. Another thing to consider is that it’s kind of already that way. We are competing against big SEO agencies that have hundreds of employees and project management software we couldn’t dream of affording. That opens them up to certain enterprise-level clients and projects at a scale that we couldn’t handle.

However, that’s fine because we don’t position ourselves that way. We position ourselves as small, boutique, hands-on, and offering one-on-one communication with clients. That’s our differentiating factor. It might end up being the same as that with AI.

We’re constantly getting clients who have left a giant SEO agency and are craving that more personal relationship. Human relationships are something that will never be replaced by AI – at least in the near term. The human relationship with your clients is going to remain no matter how much you use AI in the background to scale their projects. You can’t replace that human-on-human interaction.”

If an SEO is struggling for time, what should they stop doing right now so they can spend more time doing what you suggest in 2024?

“If you aren’t using ChatGPT for content ideation, then start doing that. Having ChatGPT come up with topic ideas, outlines, and content briefs for pieces of content will save a ton of time. I’m always going to preach that you should not have ChatGPT write the content, but the content ideation it can provide is a huge time saver.

Also, recently, we have been shifting a lot more towards reviving, revising, expanding, and improving existing content on sites that have a lot of content already. Of course, if it’s a newer site, you have to focus on creating new pieces of content and building out your encyclopedia of content on that website.

However, you might have been doing the content production thing for many years – there’s a lot of content on the website and you’re still sitting there with writer’s block wracking your brain. We have been doing SEO for one client for 10 years, and we have written about every angle of their topic that we could ever come up with. That’s an extreme example but, on any large site, there’s tons of content to work with.

Pick the pieces of content that aren’t performing, rewrite them, revamp them, make them higher quality, make them longer, and ask ChatGPT how you can make the content better. Revise and relaunch existing content instead of agonising over coming up with new topics.”

Pam Aungst Cronin is President and Founder of Pam Ann Marketing, and you can find her over at PamAnnMarketing.com.