Most advice on Google Business Profile optimization stops at the profile: fill in every category, gather reviews, add photos, keep the hours current. That work still matters. But a recovered Google Maps binary, analyzed by Search Engine Land in September 2026, made clear that the profile a business edits is only the surface of a much larger system, and that the deeper object underneath is what Google actually ranks. Optimizing the listing without understanding the entity beneath it leaves most of the opportunity on the table.
The listing is not the entity
Internally, Google represents every place as a canonical entity, assembled from many data sources and connected to its wider Knowledge Graph. The familiar Maps listing is composed later, as an interface on top of that entity. What a business owner types into Google Business Profile is not necessarily what Google keeps as the underlying record. For anyone doing this work, the entity is the more useful unit to think about. The listing is the window. The entity is the thing being ranked, and it is built from far more than the dashboard.
Why your edits keep reverting
This reframes one of the most frustrating problems in local SEO. A business updates a field in its profile, and weeks later the old value is back. The architecture explains why. Google draws on hundreds of data providers for a single place: its name from one, its category from another, its location from a third, and it runs a process to reconcile them when they disagree, weighing each source by trust and priority. An edit in the dashboard is not a command. It is one more piece of evidence entering a system that may already hold conflicting evidence from sources Google trusts more. The fix is rarely another dashboard edit. It is making the correct information consistent everywhere Google looks, so the weight of the evidence points one way.
What Google actually scores
The analysis recovered the internal ranking system for these entities, with 72 named signals, though a third of them are already marked deprecated and their exact weights were not exposed. The visible names are still instructive. Alongside familiar factors like reviews and prominence sits a cluster of behavioral signals: how often the listing is shown, how often it is opened, how many people request directions, how many click through to the website, how much query volume the brand itself generates. This is worth sitting with. A meaningful part of what feeds local ranking never touches the profile at all. It reflects what real people do when they encounter the business, meaning the listing rewards genuine demand and cannot be fully gamed by field completion alone.
A caveat matters here. Recovering the names of 72 signals is not the same as recovering the algorithm. The ranking of any given search runs through several more systems after the entity is scored, and turning the signal list into a 72-item checklist would miss how the machine actually works. The names show the categories of evidence Google can draw on, not a weighting anyone can optimize against directly.
Distance is not a fixed radius
A common mental model imagines Google drawing a circle around the searcher and ranking whatever falls inside it. The measurements in the analysis showed something more fluid. The size of the search area changed with the query and the surroundings: a common need pulled a tight radius in a dense city and a much wider one in a rural area, while a brand name pulled a far larger area than a generic category. When the researchers stripped geography from the same engine entirely, the median result distance jumped from under seven kilometers to more than four thousand kilometers, yet the underlying order of candidates barely moved. Distance shapes which businesses enter the running, not simply how the same list is sorted. Proximity still counts, but "I am closer, so I should rank higher" is an incomplete picture. Relevance and prominence decide who is even considered.
Your website is evidence about your entity
One of the more consequential findings is how tightly Maps and ordinary web search are wired together beneath the surface. Google connects web documents to place entities through a shared identifier and scores how well a given page describes a given entity. That changes what a location page or store locator is for. Its job goes beyond ranking for a query like "plumber in Boise." The page is also evidence about the entity itself, telling Google what the place is, how much of the page is genuinely about it, and how confident that connection should be. Web reputation work and local optimization are far less separate than their interfaces suggest, and a site that clearly and consistently describes the same business strengthens the entity Google is trying to build.
Reviews as structured evidence, not a star average
The semantic layer goes beyond the single category shown in a listing. Google works from a shared vocabulary of concepts, so a search for one dish connects to related cuisines and restaurant types, and inside a listing, review topics and attributes are stored as structured themes rather than raw text. This is where reviews do quiet, compounding work. Their value goes beyond the average star rating people see. Consistent, specific feedback about real aspects of the experience becomes structured evidence Google can reason over, without rereading every review each time someone asks a question. Reviews that describe what actually happens, the wait, the service, the specialty, feed the representation directly.
The AI layer raises the bar
Google is putting a conversational layer, Ask Maps and Gemini, on top of all of this, and the infrastructure to answer complex questions was largely already in place. The query is no longer just "best ramen near me." It becomes "where can six of us eat near the hotel tonight, one of us vegetarian, short wait, good recent reviews on service." To field that, Google has to know what the business is, what it offers, when it is open, what people say about specific parts of it, and whether the evidence is reliable enough to recommend. That makes completeness the emerging battleground. A business can have a polished profile and hundreds of reviews and still be poorly represented for a specific question, if Google cannot confidently connect the relevant attributes, concepts, pages, and review themes to the entity.
What to actually optimize
The practical conclusion is not a longer checklist of profile fields. The shift is in the target, from the listing to the entity Google builds behind it. Keep the profile complete, because it still matters: accurate categories, attributes, hours, photos, and a steady flow of genuine reviews. Then work the layer most businesses ignore. Make the core facts consistent across every source Google consults, so conflation resolves in your favor. Build a website that describes the same entity clearly and links its location pages to it. Earn reviews that speak to specific, repeatable aspects of the experience. Cultivate real demand and real web mentions, because prominence and behavior are doing more of the work than the dashboard admits. The question Google is increasingly asking is whether its picture of a business is complete enough to recommend with confidence. Optimizing the Business Profile answers part of that question. Optimizing the entity answers the rest.
Part of the SEO Reputation Management pillar. Related: Google Entities, Review Management, Brand Reputation Management, and the companion guide to Google Business Profile for brick-and-mortar branded search.