Your website gets cited in an AI answer. Your brand isn’t mentioned. Your competitors get the recommendations.
Would you call that a win?
That was one of the questions I discussed with Parth on SEO Talkers. We covered AI search, the GEO debate, my press release experiments and where I think businesses should spend their time and money.
My good friend, and one of the smartest SEO professionals in the game Gagan Ghotra published a detailed summary of the conversation, bringing together the main points and examples.
Read Gagan’s full summary of my SEO Talkers interview.
Here are the points I’d want a business owner or marketing team to take away.
A citation isn’t the same as a recommendation
An AI tool can link to your website without recommending your business.
It might use your article to explain a concept. It might cite your comparison page while recommending the competitors you’ve listed.
That’s why I’m cautious about case studies built around enormous percentage increases in AI citations.
Which pages earned those citations? What questions triggered them? Was your business mentioned in the answer? Were those questions relevant to someone considering your product?
A basic explainer can attract plenty of citations while doing little to help customers choose you.
I’d rather have a useful presence in a relevant buying conversation than a large citation count with no clear connection to the business.
The name matters less than the work
Call it SEO, AI SEO, GEO or AEO. My concern is what’s being done under that label.
Much of the work still involves making information accessible, improving important pages, answering customer questions and building credibility beyond your own website.
SEO has always changed. The work we do today shouldn’t look exactly like the work we did in 2017.
But change doesn’t mean we need to pretend every established practice is a new invention.
In the interview, I also talked about how AI search is bringing SEOs into more conversations with social, video and communications teams. That’s an opportunity to contribute beyond the website, provided we bring evidence and useful ideas.
My press release experiment showed why sources matter
I discussed an experiment in which I published a press release about a fictional company, including an unsupported claim about its rating.
That claim was subsequently repeated in AI answers.
The point was to demonstrate how published information can influence what these tools say. A confident answer can still trace back to a claim that nobody independently verified.
It also raises a problem with repetition. Several websites saying the same thing may look like agreement, even when they’re carrying copies of the same original release.
For anyone using AI to research a purchase, it’s a reason to inspect the sources.
For marketers, it’s a reminder that appearing in an answer and earning trust are different achievements.
Start with what AI says about your own business
Before chasing broad “best product” recommendations, check whether AI tools can describe your business accurately.
Does your homepage explain what you do? Does your About page give useful context? Are your product and service pages specific enough to answer a buyer’s questions?
Then look at the answers themselves.
If an AI tool gets something wrong, investigate where the information came from. It could be an outdated page, a third-party profile or a competitor’s comparison.
That gives you a concrete problem to address.
Publishing more content won’t necessarily fix unclear positioning or inaccurate information about your brand.
Make video part of the conversation
Video was one of the biggest opportunities I discussed.
A useful starting point is a question you already answer for customers. Explain a product, demonstrate how something works or talk through a comparison.
You don’t need a complicated production setup to begin.
The value extends beyond AI search. A useful video can support your website, sales conversations, YouTube channel and social content.
My recommendation was to bring a video proposal to the table, supported by examples relevant to your market. Give the team a reason to invest and a practical place to start.
Spend money on something worth referencing
We also discussed putting budget towards original research.
One example I shared involved using public Toronto road fatality data to create a story about changes over time. The work gave journalists something useful to cover.
That’s the kind of project I’d consider when reviewing a budget otherwise being spent on buying backlinks.
A worthwhile study can provide material for an article, media outreach, social posts and video. It gives other people a reason to reference your business because you’ve contributed something.
Measure the questions that matter
AI visibility reporting needs context.
A citation on an introductory question and a recommendation during a product comparison serve different purposes. Counting them together can hide what’s actually happening.
I’d want to know whether the business appears for relevant customer questions, how it’s described and which sources shape the answer.
Where possible, I’d connect that to enquiries, conversions and what customers tell us about how they found the business.
The goal is to understand whether the work is helping people discover, evaluate and choose you.
That was the thread running through the conversation: look past the acronym and the impressive chart. Ask what the activity is doing for the business.
Read Gagan Ghotra’s full summary of my conversation with SEO Talkers.