Benjamin Houy shut down Lorelight, a generative engine optimization (GEO) platform designed to trace model visibility in ChatGPT, Claude, and Perplexity, after concluding most manufacturers don’t want a specialised software for AI search visibility.
Houy writes that, after reviewing a whole bunch of AI solutions, the manufacturers talked about most frequently share acquainted traits: high quality content material, mentions in authoritative publications, robust repute, and real experience.
He claims:
“There’s no such factor as ‘GEO technique’ or ‘AI optimization’ separate from model constructing… The AI fashions are skilled on the identical content material that builds your model in every single place else.”
Houy explains in a blog post that clients appreciated Lorelight’s insights however typically churned as a result of the info didn’t change their techniques. In his view, customers pursued the identical fundamentals with or with out GEO dashboards.
He argues GEO monitoring makes extra sense as one sign inside broader web optimization suites reasonably than as a standalone product. He factors to examples of conventional web optimization platforms incorporating AI-style visibility indicators into present toolsets reasonably than making a separate class.
Debate Snapshot: Voices On Each Sides
Reactions present a real break up in how entrepreneurs see “AI search.”
Some web optimization professionals applauded the back-to-basics message. Others countered with circumstances the place assistant referrals seem significant.
Listed here are a few of the responses revealed up to now:
- Lily Ray: “Thanks for being trustworthy and for sharing this publicly. The business wants to listen to this loud and clear.”
 - Randall Choh: “I urge to vary. It’s a rising metric… LLM searches normally have higher search intents that result in larger conversions.”
 - Karl McCarthy: “You’re proper that high quality content material + authoritative mentions + repute is what works… That’s not a software. It’s a community.”
 - Nikki Pilkington raised consumer-fairness questions on shuttering a product and whether or not prior GEO-promotional content material must be up to date or eliminated.
 
These views seize the business stress. Some see AI search as a brand new efficiency channel price measuring. Others see the identical model indicators driving outcomes throughout web optimization, PR, and now AI assistants.
How “AI Search Visibility” Is Being Measured
As a result of assistants work in another way from internet search, measurement remains to be uneven.
Assistants floor manufacturers in two principal methods: by citing and linking sources instantly in solutions, and by guiding individuals into acquainted internet outcomes.
Referral monitoring can come via direct hyperlinks, copy-and-paste, or branded search follow-ups.
Attribution is messy as a result of not all assistants move clear referrers. Groups typically mix UTM tagging on shared hyperlinks with branded-search raise, direct-traffic spikes, and assisted-conversion experiences to triangulate “LLM affect.”
That patchwork makes case research persuasive however arduous to generalize.
Why This Issues
The principle query is whether or not AI search wants its personal optimization framework or if it primarily advantages from the identical model indicators.
If Houy is appropriate, standalone GEO instruments may solely produce partaking dashboards that seldom affect technique.
However, if the advocates are appropriate, overlooking assistant visibility may imply lacking out on worthwhile alternatives between conventional search and LLM-referred site visitors.
What’s Subsequent
It’s seemingly that web optimization platforms will proceed to fold “AI visibility” into present analytics reasonably than making a separate class.
The most secure path for companies is to proceed doing the brand-building work that assistants already reward, whereas testing assistant-specific measurements the place they’re almost definitely to repay.
Featured Picture: Roman Samborskyi/Shutterstock
Source link


                                        