It’s Go Time for GEO

We’re nearing a tipping point in AI Search from initial experimentation to robust implementation, and it’s time for savvy tech marketers to take note. 

I’ll admit to self-interest here since Firebrand sells Generative Engine Optimization services to get brands noticed in AI Search. But even if you don’t work with us, you do need to start your GEO program. Here’s why:

AI traffic is growing

Site traffic from AI sources is on the rise – and it’s highly engaged. We have seen a steady rise in AI traffic for our clients (and Firebrand’s website) since June 2025. Some saw a slight reset in November, but the numbers are climbing back again. These visitors spend 2-3x the time on site as other sources, with an engagement rate that’s significantly higher than the site average,  indicating high intent. They tend to land on mid-funnel blog posts explaining how to tackle a problem, case studies where that problem has been solved, core product and services pages, and data around the issue. Conversions from AI traffic are still low during a single session, but just this week, a client chalked up their first SAL from AI sources with an enterprise prospect. 

Buyers are actively using LLMs, primarily ChatGPT, to research B2B vendors. When they come from citations to your site, they are engaged and in the mid-funnel stage. If you are unsure of your own traffic, HubSpot now has an AI traffic report, and you can see it in GA4 using these instructions.

AI visibility tools are here

We’ve talked about the AI traffic to your site. What about brand mentions within the LLM responses themselves? To recap, the LLMs don’t share data in the same way Google does – so it needs to be generated. To do that, the AI visibility tools submit a range of prompts and analyze the responses. The prompts represent the queries your personas are likely to ask the LLMs. If you submit the same prompts over time, you can then see trends in which brands are mentioned, which are cited, and where the LLMs are getting their information. This is the baseline for your GEO efforts.

Now, there are a lot of problems with this approach. LLMs are probability engines, so how can we read anything into their responses? What if changes to responses are simply the LLMs changing their weights? How do we account for the long-term memory LLMs have on each of us that shape their responses? Since these tools use the APIs, do they get the same responses we do? What if those APIs have caches and how do those caches work? How do we actually know which prompts our audiences are even typing in?

You get the picture – there are many unknowns here. Some, like Rand Fishkin, are even calling bullshit on the whole thing. And he should know since he helped create Moz, one of the main SEO monitoring tools.

Caveat emptor – but marketers love a good graph, and what gets measured gets managed. These tools are the worst they will ever be – and at least they can help guide our GEO efforts. So what should a good AI Visibility report look like, I hear you ask? Good question – it should look like this.

Related, these tools are not free or even cheap. They use tokens per prompt per model and will poll 75-100 prompts across 3-8 models every few days or at least weekly. Then you need to analyze the results. Costs soon add up, so budget accordingly.

The tactics are solidifying

Generative Engine Optimization is a new discipline. It doesn’t have a set of repeatable tactics that specialists know will deliver consistent results, as we do with SEO. We do have plenty of early signals, however, such as adding FAQs to your site, which have measurable results like this.

Our understanding of how the LLMs return results is improving. As the GEO community shares approaches that are validated by others, we will begin consolidating on go-to strategies that should work. 

Brands that wait on the sidelines before investing will miss any early-mover advantage. That caution will minimize wasted effort but will mean they have more competition when they get going, and must catch up with their bolder competitors. AI Search fortune may favor the bold at this point. Even if AI visibility metrics are questionable, AI traffic to the site is real and measurable.

The board and the sales team will ask

SEO teams are familiar with quarterly board questions about keyword rankings for critical terms. Most get out in front of it with their reporting. Well, prepare for AI search to be high on the list of board priorities. The Wall Street Journal’s Christoper Mimms just did an excellent primer about Generative Engine Optimization (granted, with the twist that SEOs ruined the Web and are now coming for LLMs). Whose phone will ring when the board asks where the brand ranks and what the GEO strategy is? 

It’s also SKO season, and you can bet the sales team will be fully aware of how LLMs are impacting buyer behavior. They will want to know the company is doing everything it can to make AI shortlists. LLMs are now the gatekeepers for B2B buyers – so sales teams will want to make the cut. Who knows how many deals are being lost before they even begin? With a GEO program in place, that risk can be minimized.

The budget window is now

Marketing budgets are tight, and AI search is a new channel for brands. For many, there’s no ongoing GEO program, so it’s a new line item. GEO builds on the SEO program – so will need additional funding. The GEO icing still needs the SEO cake. With the new financial year, it’s time to make sure this program is properly funded and resourced. GEO is a team sport involving SEO, PR, content, web and even affiliate.

It’s go time

All the signals are there to get the GEO program going. It’s not too late. In fact, it’s still early in the race. But be sure, many brands are already off and running.