When Google Stops Sending Customers: The AEO Risk Nobody Is Measuring

1. Why did Google stop sending you customers?

Google stopped sending you customers because over 60% of dining searches now end without a click in markets where AI Overviews operates: the diner asks 'where can I eat well nearby?' and gets a synthesized answer, not ten blue links.

Classic SEO competed for a spot on that list of ten; that game is emptying out. I've seen restaurants with the same average Google position as two years ago lose 30% of organic traffic without moving a single ranking. The cause isn't a penalty: the engine no longer hands out clicks, it hands out answers. Between 2024 and 2026 zero-click crossed that 60% threshold and keeps climbing. If your dashboard still reports 'average position', you're measuring a battle that changed fields. The relevant question is no longer where you appear on the list, but whether the AI cites you inside its single answer. AEO (Answer Engine Optimization) optimizes to be cited inside the AI answer, while classic SEO optimized to win the click.

2. Classic SEO vs. AEO: what truly changes

That is the structural difference, not a cosmetic one. SEO is a battle of position: climbing one spot among ten links. AEO is a battle of authority and structure: being the source the engine picks to answer. A restaurant can sit at Google position 3 and never be cited once by Perplexity or AI Overviews, because its content isn't written in self-contained passages the AI can extract. In my work with operators I repeat it: ranking is no longer the defensible asset. The asset is the citable corpus. Optimizing only for position today is sharpening a spear for a war that already changed weapons. Those who grasp this rewrite their content so it answers full questions within the first 40 words. The metric that matters shifted from 'CTR over position' to 'citation rate per answer engine', and almost no restaurant has it instrumented. That is the real strategic problem: not the traffic drop, but that it's invisible.

3. The KPI almost no restaurant measures

Operators keep reporting 'average Google position' when the question became 'do I appear in the AI answer'. Diego F. Parra frames it this way in the boardrooms Masterestaurant advises: without citation rate on the dashboard, the leak is caught three quarters late, already turned into falling sales. To measure it, query the answer engines with the 20 or 30 real questions in your category and count how many cite you. It's a number between 0% and 100%, and most restaurants start below 10%. That percentage, not average position, is what predicts your acquisition cost a year out. Customer acquisition cost spikes when the organic channel dries up, and it does so silently because no indicator anticipates it. A restaurant that got 40% of its bookings from free organic traffic and watches that channel fall to 15% still needs those bookings: it buys them via paid ads or platform commissions.

4. Acquisition cost spikes in silence

I've measured cases where effective CAC rose from 4 to 11 dollars per diner in 18 months just from that migration. The bill arrives late and disguised as 'more marketing budget'. The point of AEO discipline is defensive: every category question where you stay cited is a booking you don't buy. That's why a citable corpus, structured reviews and brand demand work as insurance against CAC inflation. It isn't pretty traffic: it's protected margin when the organic click evaporates. Ignore it and the cost surfaces as shrinking profit, not as a traffic chart. Structured reviews and clean data are the raw material the AI extracts to cite you, and today they're a more defensible asset than any ranking. Answer engines synthesize from sources with clear structure: hours, price ranges, cuisine type, FAQ answers written in self-contained prose. A restaurant with 200 coherent reviews and a rich listing gets cited more than one with better position but scattered data.

5. Reviews and structured data: your citable raw material

At Masterestaurant we measure this by category: venues with full schema and answer-first FAQs double their citation rate in three months. The key isn't content volume, it's citability: each passage must answer a full question on its own, with the figure or fact in the first sentence. The AI won't cite paragraphs that need external context. Structure your information as if every block were to be read alone, because that's exactly how the engine reads it. The 90-day roadmap starts by measuring your baseline citation rate and ends with a defensible citable corpus, protecting your CAC before the organic channel dries up. Days 1 to 15: define the 25 real questions in your category and measure how many cite you across each engine; that percentage is your baseline. Days 16 to 45: rewrite your key content into self-contained answer-first passages and implement full schema and FAQ.

6. A 90-day roadmap to stay in the answer

Days 46 to 75: structure reviews and data, and reinforce brand demand so the diner searches you by name, not by category. Days 76 to 90: measure citation rate again and compare. A jump from 10% to 35% is realistic and means bookings you stop buying. Diego F. Parra insists: this isn't a campaign, it's installing on the dashboard the one KPI the AI era makes indispensable. Classic SEO competes for a spot in a list of ten links; AEO competes to be the source the AI cites inside a single answer. The first is a battle of position; the second, of authority and structure. The indicator that matters changed: from 'CTR over position' to 'citation rate per answer engine'. Without that KPI on the dashboard, the traffic leak is invisible until it surfaces as a sales drop three quarters later. The defensible asset is no longer the ranking, it's the corpus: a body of citable content, structured reviews and brand demand that lowers customer acquisition cost even when the organic click evaporates.