For Marketers, GEO May Feel Like Another Distraction — It's Not!

In the space of a few short years, generative engine optimisation (GEO) has gone from something theoretical to a key component of looking after your brand. With countless data points showing rapid growth of day-to-day AI use and more zero-click searches than ever, companies at every level need to understand how they show up.

Right now, though, lots of brands aren’t doing that work, in part because GEO is a perfect storm of all the things that brands hate to hear when they’re investing in marketing spend. Firstly, it’s still in the phase where it’s constantly changing. There are lots of very clever people making predictions, but the nature of AI models (and the unique ways people are using them) means there’s a huge amount of guesswork required.

Secondly, it’s very difficult to measure. So far, nobody’s cracked a reliable way to tell how your brand is showing up in the models, and more importantly, where it matters. They’re far more inconsistent than search and the way they personalise means that replicating a universal measurement framework is very tough.

And finally, in a more abstract sense, they just feel complicated. As has long been the case with SEO, there’s a sense from some marketing people that they should shy away from GEO because there’s a huge amount of technical expertise required to understand what’s going on under the surface. AI models are an even more extreme version of this, because we’re now at the point where they’re verging on black boxes – where nobody fully understands why a model does what it does.

This doesn’t need to be a blocker. You don’t need to know how an engine works to pass your driving test. You don’t need to be an expert in infectious diseases to take penicillin. You don’t need to understand how large language models are built to know that you need your brand to have a presence in them.

Thankfully, the prescription for improving LLM visibility isn't nearly as complicated as you might think.

The first step is to escape the mindset that this is a new channel to tick off the list, build the playbook, implement and move on. Treat LLMs like an audience. The most thorough, detail-oriented audience you’ll ever encounter. This is an audience that’s communicating to another on your behalf, and it doesn’t shy away from deeper research and honest opinions to do that. Your focus should be steering that communication, directly and indirectly, in the places AI learns.

But that doesn’t mean start spamming every channel you can get your hands on. Please don’t do that.

Start with the work that’s already being done and work back.

Is your PR and comms approach fine-tuned to maximise your coverage in the places AI is choosing to learn about brands like yours? If you notice AI regularly uses the same review roundups when answering customer questions about the best product, are you actively targeting those lists?

Maybe your brand is always mentioned as an option, but AI gives much more detail about your competitor. That points to AI lacking real detail about your features and attributes. Focussing on making your proposition clear in owned and earned spaces. Being mentioned alone is not a measure of success.

Or your website might be working against you now. A conversion machine that’s made for people who already know you, with SEO a long way down in the priorities. If the detail AI needs the most has been designed out along the way, you need to find a way to tell that story again.

And even where direct influence is difficult, there’s still opportunity. Wading into a platform like Reddit without a dedicated strategy to earn advocacy is dodgy at best. It’s not the place for a GEO tactic bolt on. So treat it as research opportunity. GEO is a window into many important things you may have previously turned a blind eye to. If the recurring complaint is your price point versus competitors, then you know your paid media messaging needs to address this. Focus it on why the price tag is worth it and activate it in those hard-to-reach places.

The best GEO work isn’t just about steering what AI says. It’s about being ready to act on what it exposes.

Not everything you do needs to be based on complicated data analysis. The best way to get a real understanding of how technology is impacting a brand is talk to your customers. They’ll tell you more than any estimations of prompt volumes or reports ever could.

The most important thing is to get started and be prepared to adapt. Even if there were a golden formula for success right now, it might be completely different next week. And the good news is, most of the things you can do to improve GEO will help your brand in lots of other ways too.

Becky Jenkins is organic search director at TMW, part of Accenture Song.