Nine pieces. One operating thesis.

I typed “best carbonara near me” into my phone one slow Tuesday on the floor, the way a guest two tables over had just done out loud. Google didn’t hand me a list. It handed me a paragraph — three sentences it had written itself, naming three restaurants, before the map even loaded. I knew the block. I knew the chef. The room I was standing in served a better one, and it was nowhere in those three sentences. That gap is the whole reason this batch exists.

So this is Week 1 of a new batch cadence, and it’s bigger than the ones that follow on purpose — it lays the foundation they build on. Future batches will be leaner and hook to whatever’s moving in the restaurant industry that week: a Google product update, a delivery-platform policy change, a wage-law phase-in. The foundation is this one idea. AI Overview citation is the new discovery surface, and almost no independent restaurant writes for it yet. The nine pieces below are what I’d hand the operator next to me to fix that, in the order I’d hand them over.

The through-line

Two things are happening at the same time, and they compound. One: Google’s search-result page is a paragraph the model wrote, citing two or three sources. As of March 2025 it triggered on 13.14% of US desktop searches; by spring 2026 it’s approaching one in five, and restaurant queries trigger it more often than the average. Two: the map pack still sits beneath that paragraph, and it’s still where most local-intent traffic clicks. So you’re competing on two surfaces at once — one new and shaped by paragraph quality, one old and shaped by Google Business Profile health.

Watch how fast the new surface arrived. One number tells the story, and it’s the only AI Overview figure in this whole batch with a single hard source behind it — everything after it I’ll mark as direction, not a dial.

Five of the nine pieces address those two surfaces directly. The remaining four address the operating-economics decisions that determine what your restaurant actually has to say on those surfaces — what your menu costs to put in front of a guest, who handles the discovery, what compensation model your kitchen runs on.

So before the reading order, the one test the whole wave is built around: when the model drafts that paragraph and reaches for two or three sources, is your page one it can actually lift from? Walk the three questions.

Ranking and being cited are not the same job. You can sit fourth in a list nobody reads anymore and lose to the restaurant that simply wrote the answer down.

The reading order

Clear those three questions and the order below is just the route I’d walk to get there. Each piece stands alone, but they ladder. Read them in order — the role tag on each tile tells you why the piece sits where it does. Search citation first (one through four), then social discovery (five), then operations and economics (six through nine).

What the claims in these pieces are anchored to

One more thing before you start clicking, because it’s the part I’d want to know if a stranger handed me nine articles. Every operating-floor number across the nine articles is anchored to a publicly verifiable source — platform-published commission tiers, vendor pricing pages, search-industry analyses, and Google’s own help documentation — or to an illustrative range that’s identified as such in the piece itself. Specific anchors include:

• Platform commission and fee structures. DoorDash Marketplace tiers (Basic 15% / Plus 25% / Premier 30%), Uber Eats and Grubhub published rates, payment-processing rates — all from the platforms’ own merchant documentation. Powers the DoorDash margin walk, the three-platform comparison, the 30-day playbook.
• AI Overview share of US search. The March 2025 Search Engine Land analysis (13.14% of US desktop queries) is the most-cited measurement; subsequent direction is qualitative. Powers the AI Overview citation article’s framing.
• Loyalty + email vendor pricing. Toast Rewards, Square Loyalty, Fivestars/Como, Mailchimp, Klaviyo monthly costs are taken from each vendor’s current public pricing page. Powers the loyalty comparison.
• DMV Google Business Profile audit (2023–2026). Sample of independent restaurants in the DMV; source for the four root causes of map-pack invisibility in piece #3.

Don Goldstein is a restaurant operator and runs Muntin Digital. He writes the weekly batches that ship on this page.