Generative Engine Optimisation: ranking where Google does not own the answer
What GEO is
Generative Engine Optimisation is the practice of structuring your content, your schema, and your wider digital presence so generative AI assistants recommend you when asked category questions in your space. "Recommend" can mean cited in a response, mentioned by name in an answer, or shown as a source in a Perplexity reply.
Unlike Google SEO, GEO depends heavily on signals from outside your own site. The generative engines pull from Wikipedia, reddit, industry blogs, podcasts, news coverage, and structured directories. Reputation across the open web matters more than backlinks per se.
Why GEO matters
- ChatGPT search reached ~250 million weekly users by late 2024 (OpenAI announcements). That audience is being given category answers without ever visiting Google.
- Perplexity has 15+ million monthly users and is the AI engine most likely to cite specific URLs in answers, making it the highest-traffic source of GEO referrals today.
- Claude is bundled into Anthropic's enterprise customers and a growing set of consumer apps. Lower traffic but high-influence audience.
The aggregate is a meaningful slice of the discovery layer that did not exist three years ago. Brand visibility inside these engines is now its own marketing surface, separate from search.
The engines we optimise for
- ChatGPT (with search). Cites sources inconsistently but increasingly. Pulls from Bing index plus its own crawler. Weighting unclear but strong Wikipedia presence and recent news coverage help.
- Perplexity. Cites sources explicitly with hyperlinks. Closer to a search engine in citation behaviour, friendlier to traditional SEO signals, rewards FAQ-shaped pages.
- Claude (browsing mode and through downstream apps). Limited public tooling; influence is mostly through reputation across open-web sources.
- Bing Copilot. Behaves closely to Bing search ranking factors, so traditional SEO carries over.
What we actually do
On the $800 per month AEO/GEO add-on, four streams of work run alongside the regular SEO retainer.
- Page-level content shape. Same AEO-shaping techniques (answer- first leads, FAQ schema, citable statistics) apply. Plus engine-specific tweaks: tighter section H2/H3 hierarchy for Perplexity citation, more aggressive entity markup (schema.org) for Claude crawling.
- Third-party citation building. Pitch list of AU industry blogs, podcast appearances, expert-quote opportunities (HARO, Qwoted, SourceBottle). Goal: build a footprint that generative engines pull from when forming category recommendations.
- Wikipedia and structured directories. If the brand is notable enough, a Wikipedia page substantially lifts citation rate in ChatGPT and Claude. Where not notable enough, structured directory presence (industry databases, AU government registers) plays a similar role.
- Prompt testing and coverage tracking. Quarterly snapshot of how target category prompts resolve in ChatGPT, Perplexity, Claude. The output is a matrix showing which engines mention you, how, and how often.
How we measure GEO performance
GEO measurement is less mature than SEO measurement. We do not pretend otherwise. Three things we do measure, monthly or quarterly:
- Prompt coverage. Defined set of category prompts, run quarterly, scored on whether each engine mentions you. Trend tracked over time.
- Referrer traffic. GA4 referrer filtering picks up Perplexity and ChatGPT visits. Volume is small but growing, and the engagement is high because the user already read about you before clicking.
- Branded query lift. GEO citations drive branded search later (someone hears your name in a Perplexity answer, then Googles you). Branded query volume on tools like Google Search Console is a lagging but reliable indicator.
The opinion we hold strongest on GEO
GEO is the most underpriced marketing channel right now. Most agencies do not offer it. Most clients do not ask for it. The cost of building a credible Perplexity and ChatGPT presence over six months is materially less than the cost of buying equivalent paid search traffic, and the asset compounds. The window where this is cheap will not stay open forever.
Frequently asked
How is GEO different from AEO?
AEO targets Google's AI Overviews specifically and follows Google's ranking patterns closely. GEO targets ChatGPT, Perplexity, Claude, Bing Copilot, and other generative engines that crawl differently and weight different signals (third-party citations, Wikipedia, podcast mentions, reddit). Both matter; they overlap but are not the same job.
Does GEO work?
Yes, with caveats. We can move ChatGPT and Perplexity recommendations on category queries with deliberate work. Claude is harder to influence (limited tooling). Bing Copilot tracks closely with regular Bing SEO. The improvements are not instantly measurable like Google rankings, which is why we run prompt-testing as quarterly snapshots.
What does a GEO retainer actually deliver?
On the $800/mo AEO/GEO add-on: monthly prompt-coverage report across ChatGPT, Perplexity, Claude. Pitch list for high-citation third-party sites (industry blogs, AU-specific Wikipedia categories, podcast appearances). Content shape review on every page. Schema enrichment for engines that pull structured data.