How Google picks who to show

Chapter 1 argued that the map is where a local customer decides. The next question follows on its own: how does a business get onto that map at all, and why do some businesses appear at the top while others sit several screens down?

Google answers part of this in public, and the public part is worth reading closely, because it is the only part anyone outside Google can rely on.

Three things Google names

Google says its local results are ranked on three things: relevance, distance and prominence.

Relevance is how well a business matches the search. It comes from the category, the services the business lists, the words on its profile and its website, and its photos. A jeweler who resizes rings every day but never says so anywhere is a weaker answer for ring resizing than a jeweler who says it plainly, even if the work itself is identical.

Distance is how far the business is from the person searching, or from the place named in the search. This is the part that changes from block to block, and we will come back to it.

Prominence is how well known a business is. Google describes it as drawing on things like reviews and ratings, how often the business is mentioned and linked to elsewhere on the web, and its website. A business that many people have dealt with and written about is a safer answer for Google to give a stranger.

Google does not publish the details: no weights, formulas or list of factors in order of importance. Anyone who tells you they know exactly how the three are combined is guessing, and anyone who sells a method based on that guess is selling the guess.

The three also explain most of what an owner finds puzzling about the map. A business can match a search perfectly and still sit fifth, because the two above it are closer. It can be the nearest shop and still sit fifth, because the others are better known. Reading a result this way, as a mix of three separate things, is more useful than reading it as a verdict on the business.

Intent comes before ranking

Before Google can rank anything, it has to work out what kind of question it is answering.

A person who types ring resizing may want to read about the job, buy a kit and try it at home, or find a shop to walk into this afternoon. Three different jobs, three different answers. The ranking only starts once Google has decided which one this is.

It reads intent from many signals at once: the words of the search, where the phone is, what the person searched before, the time of day. Nobody reads intent from a few words as well as Google does. The map at the top of the results page is its verdict. When the map is there, Google has decided the person wants to go somewhere.

The words near me are a hint, not a requirement. A person standing in Bay Ridge who types ring resizing gets a map of jewelers around them without typing the words at all. When Google's suggestions offer to add near me to a service, that tells you many people who search for it want a place nearby, which is useful to know, but Google did not need the phrase to work it out.

Because Google now reads much as a person does, the job of a page has changed from fooling a machine to making yourself understood; chapter 7 takes that up. Longer searches state intent more precisely, which is exactly why they are narrower. Jewelry store could mean buying, selling, repairing or browsing. Broken necklace chain repair means one job, and only businesses that clearly do that job are a good answer. Chapter 4 takes that idea apart with our own numbers, and it is the hinge the whole method turns on.

Distance, in practice

Most people never type a neighborhood. They type the service, or the service and near me, and Google takes the place from the phone. So the map is drawn around the customer, not around the business.

This is why the map differs every few blocks. In South Brooklyn our measurements find 8 separate micro-markets for med spas and 7 for jewelers. Two customers a few blocks apart can see different first places for the same search, and both are being given a correct answer to the question they asked.

The practical consequence is uncomfortable for owners. The area around your door where Google is willing to show you is usually smaller than you hope. An owner who thinks of the whole borough as their territory will read a result from three miles away as bad news, when it is really a different market with different leaders.

It also means the neighborhood name in a search matters less than most owners assume, and mostly when the customer is looking for somewhere other than where they stand: planning a visit near work, or asking on behalf of someone else. Chapter 6 builds its definition of a seed around this.

Prominence compounds

Prominence is the slow one, and the one that decides the big searches.

The favorite on a head term has been collecting attention for years. Being shown first brings it more views and more customers. More customers write more reviews, and reviews are part of what prominence is made of. The advantage compounds quietly, in the background, while everyone else is open for business.

The scale of that lead in our area: the median med spa has 51 reviews and the top tenth have 285 or more. The median jeweler has 18 and the top tenth have 227 or more. Those are not numbers that arrive in a good month. They are what years of steady trade look like when it is written down on a profile.

Prominence is a description of what has already happened, which is why it is slow to gain and slow to lose.

For an owner, that cuts two ways. A big search with a long-standing leader is not a fair fight, and losing it is not evidence that you are doing something wrong. A smaller search with no established leader is a fair fight, and that is where the method in Part II sends you.

What an owner can move

The useful question about the three is which one you can do something about.

Relevance. This is the field you control most directly. Set the right primary category. List the services you actually provide, in the words customers use. Describe the work in plain language on the profile and on a page of your own. Show photos of the work itself. None of it is clever, and all of it is within reach of a small team in an afternoon.

Prominence. You cannot manufacture a reputation, but you can earn one in public. Ask customers for reviews over time rather than in a burst. Answer them. Be mentioned in the places your neighborhood reads: a local paper, a community group, a sponsor's page, a supplier's list of stockists. Keep your website current, since it is one of the things Google reads when it judges how well known a business is.

Distance. You cannot move the shop, and you should not try to fake a second location. What you can do is keep the address and the service area accurate, so Google is measuring from the right door. A wrong pin on the map, or a service area drawn to cover a borough you do not really serve, does not buy you reach. It gives customers a reason to arrive somewhere you are not.

In practice the work runs in that order. Relevance is the fast one: it responds to an afternoon of editing. Prominence is the slow one: it responds to a year of ordinary good trade, written down in public. Distance is fixed, and the honest response to it is to compete where you actually are.

One item sits before all three. A profile that has never been claimed can still be read by customers, but nobody is looking after it, and nobody is answering what is on it. That is the cheapest problem in this book to fix, and chapter 3 starts there.

When the question goes to an AI assistant

More and more people will not type a search at all. They will ask an AI assistant on their phone, often out loud: where can I get my dog groomed this Sunday, he's big and scared of clippers, somewhere near Bay Ridge. The assistant answers with a recommendation or two and a sentence about why.

Where does it get that answer? From the same places this book is about. Google's own assistant reads Google Maps and its reviews directly. Other assistants draw on their own map sources and directories, and all of them read the same kinds of material: business profiles on the maps, reviews, the business's own website, local news. An assistant has no way to visit your shop. It knows what is written about you in public, and nothing else.

Three things follow, and none of them changes the method in this book. Each makes it matter more.

The questions get longer. A spoken request is a long tail in its purest form: one service, one problem, one place, one day. The assistant matches the meaning of the request against what it can read, so a profile and a page that describe the work in plain detail are what it can match.

The window gets narrower. A map shows three places. An assistant often names one or two. Being the obvious answer to a small, specific question becomes worth even more.

Reviews become the reason. When an assistant explains its choice, the explanation usually comes from what customers wrote: good with nervous dogs, fixed my clasp while I waited. Reviews that say in the customer's own words what was done and how it went are the raw material of that sentence.

One practical point. Because different assistants read different map sources, the basic facts about the business, its name, address, hours and services, should be the same everywhere they appear, not only on Google.

Where owners usually go wrong

These four mistakes come up again and again, in every trade we measure.

Believing in a secret trick. No setting, hidden field or trick of wording moves a business to the top. The people selling one are selling certainty about a system whose details Google keeps to itself.

Paying for reviews. Bought reviews are against the platform's rules, they are often obvious to customers, and they put the profile at risk. A steady trickle from real customers is worth more, and it is the only kind you can defend if anyone asks.

Stuffing keywords into the name. Words in the name do seem to count on the map, which is exactly why the trick is tempting. Adding words that are not on the sign breaks Google's guidelines and risks the profile; a name that genuinely says what you do is the honest version of the same idea (chapter 3 explains both).

Judging from your own phone, at your own desk. Your phone knows where you have been and what you have searched, and your desk is one point on a map that changes every few blocks. A single reading from one spot is an anecdote. Chapter 8 replaces it with a monthly check from a few fixed places.

All four are what a reasonable owner does when nobody has explained how the map works.