I have skimmed over a lot of detail - so give the full piece a read - but this is a valuable approach. Thanks, Amaka! www.linkedin.com/pulse/how-i-...
🧵 6/6
Posts by Chris Green SEO
There is one small health warning, though - those who need a one-sized-fits-all-silverbullet-all-in-one-solution possibly don’t have time to sit, think deeply and approach this challenge in a thoughtful way as Amaka does.
🧵 5/6
“Do you need the perfect prompt? Hell no! You need a clear head, a sense of what matters, and a willingness to push the AI in the direction you need it to go.”
🧵 4/6
Some are good IF you take the time to tweak/customise and enhance, but I think that’s where the principles-first method really stands out.
As well as delivering something more often workable, I think this also makes AI-assisted work more approachable.
🧵 3/6
LinkedIN is wall-to-wall “comment ____ for my 20 prompts that DO EVERYTHING” or similar. Spoilers - they are almost always engagement bait and don’t live up to their promise.
🧵 2/6
I enjoyed this one. What happens when we abandon “do-all” processes and instead adopt a series of principles and apply them with our own experience/creativity?
Amaka - who I had a blast working with on the 2025 Web Almanac - nails it when she says:
“principles, not fixed steps.”
🧵 1/6
This isn’t a score that has the potential to better represent LLM’s understanding of your brand within the training data, rather than a more simplified “how often does the search tool return my URLs?”. Could be super powerful stuff! waikay.io/ai-topical-p...
🧵 7/7
These three are combined to establish a single score, but one with much more definition (see attached).
I want to dig into this more - my questions around the dataset and how you establish something that is representative of “good” coverage for topics and competitors.
🧵 6/7
- Topic breadth - how many different topics you’re associated with - from strong connections to signs of emerging associations
- Topic Concentration - are you really strongly associated with one or two topics, or do you have a good distribution across a number of topics?
🧵 5/7
Rather than a raw visibility %, this score is composed of:
- Topic depth - how strongly you’re associated with key topics in your niche when compared against the strongest brand in the space
🧵 4/7
If you can agree/trust that your data set is representative and fair, then it still has value. But it could be much more useful.
Waikay published a blog post on their response to this in the AI space, “Topical Presence,” which immediately struck me as something worth investigating further.
🧵 3/7
What’s not to like?
Visibility scores have some significant limitations:
- Shaped entirely by the keyword/prompt selection - smaller data sets suffer worse here
- Visibility does not equal demand
- Without segmentation - niche, intent stage, persona, etc. - visibility can be meaningless
🧵 2/7
Visibility scores have always been a bit hit-or-miss for me.
On the face of it, a % score which expresses how visible you are (relative to what's possible), is a useful stat. Whether for “traditional” rank tracking or prompt tracking, it gives us a simple way to show how well we’re doing.
🧵 1/7
(for example, buff.ly/WCkdsoH)
I think we should embrace these tools and improve these standards if we want to level up SEO as a whole.
📸 5/5
Whilst this isn’t a problem, it does show something really clear - the power/influence these tools have over our overall SEO strategy. If you’ve been in the industry for 10 years or more, you’ll remember rogue updates which can introduce bugs to millions of websites too
🧵 4/5
But this includes other tendencies, like the use of Meta Robots Follow/index - these are tags that are not needed (implied if no other directives are present). We could see in the data that SEO plugins/tools were almost entirely responsible for this.
🧵 3/5
This can be seen in useful ways; for example, you can see that structured data growth is highly correlated with SEO plugin use, compared to those who don't use it.
🧵 2/5
What are the forces shaping SEO most profoundly?
This is something I spent a lot of time digging into as part of my work on the SEO Chapter of the Web Almanac.
buff.ly/fBtlcg7
🧵 1/5
I would be tracking thousands of keywords 8 times a day, long into the future - that is wasteful, and I think we’ll see diminishing returns.
This data, plus the CTR curve from Google Search Console, is useful for building a truer picture of what is going on, but it does take work.
🧵 7/7
I’m testing at a 4-hourly cadence here, too. Daily just is not granular enough to be useful. What is most important is that this testing is likely best used as indicative rather than definitive.
🧵 6/7
Together, I can do a few things with them:
- Establish which metric correlates best with actual clicks
- Create a view of what SERP makeup is less hospitable for organic clicks
- Build a visibility metric that combines the above to make a more insightful at-a-glance stat.
🧵 5/7
In this instance, I wanted to see Pixel Depth, Position (absolute), and Organic Position all together, plus a list of the features above. You can see in the screenshot that none of those metrics in isolation are very helpful because they don’t capture the context of the SERP well enough.
🧵 4/7
It’s hard to see the exact relationship between position (overall), organic position and how far a user has to scroll to see your result. Tools, like Advanced Web Ranking, do a great job at this, but I wanted to test out building my own process and running it on Cloud Run and Data for SEO.
🧵 3/7
But something Aleyda raised recently is that, from the top-level data, it’s not AIOs that are often driving the worst (in ecom), but paid search and other non-AIO features.
Off the back of this data, I revisited something I used to use more of - pixel depth tracking.
🧵 2/7
Are AIOs the main reason you’re losing clicks?
AI overviews are big and potentially impact clicks - that’s widely known and feared, and some industries are suffering heavily from this already.
🧵 1/7
I’m on a panel on GEO and AI Search at Similarweb on April 28.
Expect insights from Ethan Smith @dixonjones.bsky.social @sagibson.co.uk @chris-green.net and Rice Tong. Drinks, networking, and the future of search. 🍻
Free tickets www.tickettailor.com/events/cjeve...
This link implies that most search tools in AI chatbots are based on Google results directly or indirectly, which is fairly well known now. This won’t last forever, but for now, your “GEO actions” have SEO consequences.
Great study, Lily! lilyraynyc.substack.com/p/are-citati...
🧵 4/4
"fundamentally connected to SEO performance: if you drop in organic search, you can likely expect a corresponding drop in citations not only from Google’s own AI search products, but from other LLMs like ChatGPT, which appear to also be heavily reliant on Google’s search results.”
🧵 3/4
This blog from Lily Ray is one that came out over a month ago, but I think the data is valuable, the conclusions are worth your time, and the link she is describing is one we must monitor closely, as it will likely not remain fixed.
“We have even more evidence that AI search is...
🧵 2/4
Links in visibility in “traditional” search and “AI search” keep coming thick and fast. Meaning, you can’t safely invest in SEO without impacting GEO/AEO or the other way around.
We should be waaaay beyond arguing this by now, but it’s just not going away.
🧵 1/4