A vendor calls your restaurant and offers to submit you to two hundred directories for ninety-nine dollars a month. The pitch sounds reasonable. More places you appear, more people find you.
Decline it. Two hundred submissions solves a problem you do not have, and creates one you do: a hundred and ninety of those records will contain whatever the vendor typed, and if any of it is slightly wrong you now have a hundred and ninety wrong records to chase.
Discovery for restaurants concentrates heavily. A short list of platforms produces nearly all of the calls and visits that come through listings, a small set of aggregators quietly feeds everything else, and the remaining tail exists mostly to be sold to you.
The four that carry it
For most independent restaurants in the United States, real listing traffic comes from:
- Google Business Profile, which is the majority of it by a wide margin and the only one worth checking monthly
- Apple Business Connect, which drives navigation and Siri results for every iPhone user in your neighborhood
- Yelp, which still sends meaningful call volume in some markets and almost none in others, worth ten minutes to determine which yours is
- Bing Places, which is cheap to maintain and feeds a set of downstream surfaces including some assistant and browser integrations
Get those four correct and identical and you have captured most of the available return. Google in particular deserves separate attention because it is the listing most likely to be producing calls you never counted, which is the subject of Google Business Profile phone optimization.
Note that Yelp is the one where the phone number displayed may not be yours. Directory platforms have used call-forwarding numbers to attribute calls to themselves, and the consequences for your call data are laid out in Yelp's number substitution. Check yours before assuming it matches.
The layer nobody looks at
Underneath the platforms sit data aggregators, which compile business records and license them onward to navigation apps, in-car systems, smaller directories, and anything else that needs a database of businesses without building one.
You will probably never visit their websites. Their records still reach customers. An address from a location you left in 2019, or a phone number you stopped using two providers ago, can sit in an aggregator and keep repopulating listings you have already corrected. Operators who report that a fix "didn't stick" are usually describing this.
The tell that you have an aggregator problem is a wrong fact that keeps coming back. You corrected the address on a map app in March, it was right in April, and by June it has reverted to the old suite number. Nobody is undoing your work. The app is refreshing from a licensed record you never touched.
Corrections at that layer are slow. Submit, note the date, and check back in a month rather than resubmitting weekly. Some require a phone call. It is genuinely tedious and it is also the difference between fixing the symptom and fixing the source, which is the same underlying problem as being recognized as one business rather than several.
What the long tail is for
There is a legitimate case for a handful of listings beyond the main four, and it is narrower than the pitch.
Cuisine-specific and community directories can send real traffic when they have an actual audience in your city. A regional food blog's restaurant index, a neighborhood business association, a university's off-campus dining list, a local tourism board. These are worth doing because a human maintains them and humans use them, not because of any effect on ranking.
The test is simple: can you name a person who would plausibly use that site to pick a restaurant? If not, the listing is not going to do anything for you, and the marginal value of an identical copy of your data on a site nobody visits is close to zero.
There is one exception worth making, and it is chambers of commerce and neighborhood business associations. Those listings tend to be maintained by a human who will actually correct an error when you email them, and they occasionally get picked up by local press and city guides looking for a source. They cost a membership you may already have and about fifteen minutes. Treat them as a small, cheap addition rather than as a category worth systematizing.
Delivery marketplace listings are a separate decision entirely, since those come with a commission structure rather than a submission form. The arithmetic there sits in voice AI versus delivery app commissions.
Consistency beats count, and it isn't close
The value of a citation is corroboration. Platforms use agreement across sources as evidence that a business is real, at that address, reachable at that number. Ten sources saying the same thing is useful. Ten sources saying nine different things is worse than three sources agreeing, because contradictions make a system less confident rather than more.
So the sequence is: fix what exists, remove duplicates, then consider adding. Never the other way around. Punctuation and abbreviation count, which sounds pedantic until two records fail to merge because one says Suite 4 and the other says #4. NAP consistency covers the specific fields and the ways they drift.
The phone number is the field to be strictest about. It is the attribute platforms weight most heavily when deciding whether two records describe the same business, and it is the one that changes when you switch providers and forget which listings carry the old one.
The quarterly check
Ninety minutes, four times a year, and it prevents nearly everything above.
- Open your four main listings signed out on a phone and confirm name, address, phone, hours, and categories
- Call the number shown on each one and confirm it rings your restaurant and gets answered
- Search your name plus each old address, old number, and former trading name, and report any duplicate that surfaces
- Update holiday hours for the coming quarter on every listing, since a wrong holiday entry costs a full day
- Check that any listing showing a menu is showing the current one
The second item on that list is the one people skip and the one that costs money. A listing with a perfectly correct phone number attached to a line that rings out during your dinner rush is not a working listing. Every directory you maintain is ultimately a machine for producing phone calls, and the value it returns is capped by what happens when those calls arrive. Before you spend another hour on listings, spend one on counting how many of those calls you miss.
If that number is small, keep tidying directories. If it isn't, the directories are already doing their job and the phone is where the money is going.