AI search KPIs: Focus on inclusion, not position

We need to have a talk about KPIs and AI search.

I’ve observed numerous SEO professionals on LinkedIn and at conferences talking about “ranking No. 1 on ChatGPT” as if it’s the equivalent of a No. 1 ranking on Google:

On Google, being the first result is often a golden ticket.

Going from No. 2 to No. 1 in Google search will often result in 100%-300% increases in traffic and conversions.

This is almost certainly not the case with AI responses – even if they weren’t constantly changing.

Our team’s research shows AI users consider an average of 3.7 businesses before deciding who to contact.

Being the first result in that list on ChatGPT isn’t the golden ticket it is in Google search.

This being the case, the focus of AI search really should be on “inclusion in the consideration set” – not necessarily being “the first mentioned in that set” – as well as crafting what AI is saying about us.

User behavior on AI platforms differs from Google search

Over the past several months, my team has spent more than 100 hours observing people use ChatGPT and Google’s AI Mode to find services.

One thing came into focus within the first dozen or so sessions: User behavior on AI platforms differs from Google search in ways that extend far beyond using “natural language” and having conversations versus performing keyword searches.

Which is overstated, by the way. About 75% of the sessions we observed included “keyword searches.”

One key difference: Users consider more businesses in AI responses than in organic search.

It makes sense — it’s much easier to compare multiple options in a chat window than to click through three to five search results and visit each site.

Dig deeper: From searching to delegating: Adapting to AI-first search behavior

AI users don’t stop at the first result

In both Google AI Mode and ChatGPT, users considered an average of 3.7 businesses from the results.

This has strong implications for the No. 1 result – as well as No. 4.

The value of appearing first drops sharply — and the value of appearing lower rises — when, in 75% of sessions, users also consider businesses in Positions 2 to 8.

What’s driving conversions isn’t your position in that list.

---

Why do businesses with lower rankings end up in the consideration set in LLMs?

First of all, these aren’t rankings.

They are a list of recommendations that will likely get shuffled, reformatted from a list to a table, and completely changed, given the probabilistic nature of AI.

That aside, AI chat makes it much easier to scan and consider more options than Google search does.

Let’s look at the Google search results for “fractional CMO.”

If a user wants to evaluate multiple fractional CMO options for their startup, it’s more work to do so in Google Search than in ChatGPT.

Only two options appear above the fold, and each requires a click-through to read their website content.

Contrast this with the experience on ChatGPT.

The model gave them eight options, along with information about each one.

It’s easy to read all eight blurbs and decide whom to explore further.

Which leads to the other thing we really need to focus on: what the model is saying about you.

A bigger driver than being first on ChatGPT: Being a good fit

Many search marketers focus on rankings and traffic, but rarely on messaging and positioning.

This needs to change.

In the case of the response for an ophthalmologist in southern New Jersey, you get an easily scannable list:

Roughly 60% make their entire decision based on the response, without visiting the website or switching to Google, according to our study.

So how do you drive conversion?

Deliver the right message — and make sure the model shares it.

Dr. Lanciano may be the best glaucoma specialist in the area. But if the model highlights Ravi D. Goel and Bannett Eye Centers for glaucoma care, and that’s what the user needs, they’ll go there.

Bannett Eye Centers appears last in the AI response but may still win the conversion because of what the model says about it — something that rarely happens in Google Search.

Visibility doesn’t pay the bills. Conversions do. And conversions don’t happen when customers think someone else is a better fit.

Dig deeper: How to measure your AI search brand visibility and prove business impact

As SEOs shift toward Dig deeper: , a mindset shift needs to occur

We’re still thinking about AI search the way we’ve thought about SEO.

In SEO, the top result captures most of the traffic. In AI search, it doesn’t.

AI users consider more available options.

Responses — and their format — change dramatically with each request.

“Winning” in AI search means getting into the consideration set and being presented compellingly.

It’s not about being first on a list, especially if what’s said about you misses the mark.

In other words, SEOs who think like copywriters and salespeople will drive outcomes for their organizations.

Dig deeper: Is SEO a brand channel or a performance channel? Now it’s both

View the full article