There is a scene in Project Hail Mary where Ryan Gosling climbs into the pilot’s chair and the AI running the ship assumes he is the pilot. It announces that it is handing control over to him.
He immediately jumps out of the seat because he has no idea how to fly the ship.
I see much the same expression whenever I start showing clients the tools we use to improve a brand’s visibility in large language models and suggest they use them too.
They are having what I call a Project Hail Mary Moment.
A PHMM, as nobody in the world says except me, and perhaps now you.
I understand the reaction.
The world of AI search is murky. There are no definitive ways to guarantee success, although throwing a million dollars at the problem probably does not hurt. There are no universally accepted ranking metrics, measurement standards or, in many cases, even agreed terminology.
There is not always agreement that anything different needs to be done at all.
Google says nothing special is required beyond good SEO. That may be broadly true in a Google-only world, although my own experiments suggest the reality is more complicated.
Big brands appear to be moving quickly, but much of what I see still looks tactical rather than strategic.
Try this simple test:
“I am looking for a new pair of running shoes.”
Do not specify your location, budget, running style or preferred brand.
Type exactly the same prompt into ChatGPT, Gemini and Perplexity. Then try it in Google Search.
The answers will usually be very different. Different brands appear, different sources are used, and each platform makes its own assumptions about what the person actually wants.
To me, that suggests we are often seeing the result of disconnected tactics rather than a clear long-term strategy.
Tactical work matters. AI platforms can pick up and reuse new information quickly enough that some changes may show results within a few weeks. That makes short-term optimisation tempting.
But tactics should support strategy, not replace it.
The more important job is to decide how a brand wants to be understood, positioned and trusted over the long term.
That means making sure the brand is consistently associated with the right subjects, supported by credible evidence, and recognised as an authority across its own website and the wider web.
The brands that do well in AI search will not necessarily be the ones chasing every new trick.
They will be the ones making it easier for AI platforms to understand who they are, what they know and why they are worth citing.
Preferably without giving their clients a Project Hail Mary Moment.
#GenerativeEngineOptimisation #AISearch #GEO #BrandStrategy #DigitalMarketing
About the Author
Simon Royle is a Generative Engine Optimisation specialist, author and cofounder of GeoAnalyzer Pro.
His work focuses on AI-search visibility, entity clarity, citation analysis, website trust, content structure, supporting technical SEO, competitive analysis and practical GEO measurement.
Simon is the author of The GEO Success Series:
- How to Build an AI-Ready Website
- How to Build a GEO Trust Bank
- How to Measure Your Visibility in AI Search
His approach is practical and evidence-led, with an emphasis on helping organisations become easier to find, easier to understand, safer to recommend and more deserving of citation.
Learn more at Simon-Royle.com.