A few weeks back I attended a marketing meeting where tools to measure chatbot visibility and referrals was shared. I watched with curiosity to see the traffic figures of major AI tools to be in the low hundreds. In the case of Perplexity, it had reached double figures. This was a very large company too.
I was honestly mystified - was I missing something? On LinkedIn, chatbot visibility and referrals are nearly constantly being described as the new frontier and chomping into Google’s market share. SEO agencies are repositioning themselves as ‘GEO’ (Generative Engine Optimisation) or ‘AEO’ (Answer Engine Optimisation), along with a host of other acronyms. Software startups measuring AI visibility are quickly turning into an overgrown jungle.
Yet here I was, in a virtual room, staring at numbers that I wouldn’t bat an eyelid at. The referrals were so small that I might as well be tracking anything - the local newspaper website could easily refer just as many as this tiny stream.
It only served to harden my view that despite ChatGPT’s growth this year, GEO and AEO are not really things. There’s even less difference between them and SEO as there was with ‘Inbound Marketing’ being wheeled out by Hubspot (and evangelised by Rand Fishkin) some 15 years ago. To be fair to Dharmesh, Brian and Rand, I actually quite liked Inbound Marketing - because it grouped content based disciplines together. GEO, meanwhile, is a fairly daft division.
What is it about SEO and nomenclature? I can only assume that many in SEO are uncomfortable with it being a historically geeky domain, or associated with snake oil and spam, to the extent that they clamour to reinvent it every so often. I’ve got to admit I’ve never enjoyed the SEO label, because SEO really isn’t one thing at all.
But everything I see about GEO is SEO. There is absolutely no point of difference in how to appear. Meanwhile, the numbers are really insignificant. I wrote about this in AI in SEO: New Forms of Competition published in March. It really feels like the chat around GEO has lost touch with reality at times.
To remind myself of GEO’s flaws, I wrote a relatively spicy LinkedIn post (at least for me) as follows:
10 reasons why 'GEO' (the AI-SEO acronym) is not a thing.
AI Overviews account have visibility for just under 12% of Google searches.
That's pretty big, but it's still Google.
AIO affects long tail, low value, information searches.
Perplexity is utterly insignificant as a search proposition. Under 30m monthly users is microscopic - so much so that it has abandoned its ad play.
ChatGPT hardly sends any traffic as a %. It's most commonly well under 2%.
Are you specifically optimising for Bing? Why not? It sends more traffic than ChatGPT.
Visibility is not enough. It's also not very trackable - and it's certainly not good focus to be optimising to 'appear' in an LLM.
Every graph I see on LLM referal, every post or presentation, shows miniscule sessions in both absolute and %. Social media is way bigger.
There is absolutely no difference between SEO and GEO in terms of what you should do to rank/appear.
As far as I can tell, GEO, AIO, ETC is a bubble created by software companies as a new 'growth' area (it's still tiny) and agencies jockying for relevance.
I cannot believe the amount of posts, tools and talks about LLM tracking.
It's bizarre. It all flies rapidly away from the facts.
Impressively I had several comments pointing out ‘new ways’ of creating content that would help improve visibility on chatbots. There included ‘best lists’ and ‘comparison tables’ (both of which are absolutely standard in SEO). Another was: now there’s an added layer: making your content structured, trusted, and machine-readable… so SEO?
I then of course had a fairly tedious section mentioning old fashioned thing. I quote: Your take is like advocating that the buggy whip manufacturer shouldn't worry about horseless carriages because everyone has a horse. I didn’t use the word ‘worry’ so this is quite the assumption. Then another This reads like someone on a typewriter complaining tht PCs will never catch on. And all these lofty claims just because I said GEO wasn’t a thing. Another point: No one can predict the future.
On the flipside, Rand Fishkin of Inbound Marketing fame did comment in a somewhat abstract way, but I think he was suggesting that GEO was a Golden Goose. ChatGPT, ironically, didn’t get it.
My old chum, PR Prince in the North, James Crawford wrote an amusing riposte, which I must say had some validity and is quoted here:
I strongly disagree.
If you measure the value of visibility in LLMs on traffic by referral then I agree. But visibility in LLMs brings with it the benefits of old school brand metrics like understanding and consideration as well as pure visibility. That outcome might not be as conversion ready as traffic from Google but it’s got huge benefits to reputation and other classical PR and brand metrics
In many ways, this is the most perfect deployment of ‘I strongly disagree’ I’ve ever seen. And the other great JC’s comment got a whole 7 likes.
But I’m still left wondering if there are any concrete figures behind this new form of visibility, or is it another balloon of hot air?
And for once, most of the comments agreed with me. The people of SEO have spoken. GEO is not a thing.
