Yelp counts a bookmark as a lead. Judge it by the customers who paid

Yelp counts a customer lead whenever somebody visits your page and does almost anything — calls, taps for directions, clicks to your website, bookmarks you — and bills ad clicks. Neither is a customer. Record each Yelp call and quote request with the customer's real phone number, match those to closed sales, and judge Yelp on cost per paying customer.

A plumbing company in a mid-sized city opens its Yelp dashboard at the end of the quarter and sees 420 customer leads for $3,600 of advertising. That is $8.57 a lead, which is cheaper than anything else it buys. The owner renews. Nobody in the building could say how many of those 420 leads paid an invoice, because the dashboard does not know and the invoices are somewhere else.

This is not a post about Yelp being dishonest. Its documentation is unusually clear about what it counts, and once you read it, the problem is obvious: a Yelp "lead" is a much wider thing than the word suggests, and the contact details you need to connect a lead to a sale are, by design, partly hidden from you. Both can be worked around. Neither is fixed by looking harder at the dashboard.

What Yelp counts as a lead

Yelp defines a customer lead as a page visit followed by any action on the page: a call, a website click, a direction or map view, a check-in, a bookmark, a photo upload or a message. It splits them into ads leads, where the person clicked your ad first, and organic leads, found in Yelp's ordinary search results.

Read the list again with a sales ledger in mind. A bookmark is somebody saving you for later. A map view is somebody checking how far away you are. A photo upload is often an existing customer adding to a review. All of these are real and mildly encouraging, and none of them comes with anything you could use to find that person among your customers. The deduplication rule is sensible, but it works per action type, so one person who views your map, clicks your website and calls you within ten minutes can be three leads.

A Yelp customer lead is a measure of attention on your page. Only two kinds of it — calls and quote requests — can ever become a record you can follow to a sale, and Yelp gives you a partial version of each.

What you actually pay for

Yelp Ads are sold mainly on cost per click, priced in an automated auction. Yelp's advertising terms define a click broadly: tapping the ad, but also requesting information, a quote, a service, a reservation or an appointment from it. The bill is for interactions with the ad, decided by Yelp, not for customers.

None of that is unusual for an ad platform; Google and Meta also decide what counts and price it at auction. Yelp's help pages add two reasonable protections: it does not bill for clicks that arrive in rapid succession from the same user, and you will never exceed the monthly budget you set. The point is narrower. The platform's billing unit and its reporting unit are both events on Yelp, and the only unit that matters to you — a customer who paid — is recorded in neither.

For an advertiser, that is the arithmetic of a rising price. If your cost per click goes up 10% and nothing else changes, your cost per customer goes up 10% too. Whether you can absorb it depends on how many of your Yelp leads close and what they are worth, and that is the number this post is about.

Calls: Yelp shows you part of the number

Yelp's call reporting, free to advertisers, replaces the number on your page with forwarding numbers that ring your main line, and lists each call's date, duration and a partial caller number. A partial number cannot be matched to a customer, so the full number has to come from your own records.

The partial number is a privacy choice and a fair one; Yelp's advertising terms also ask businesses using its call tracking to limit access to call recordings and not to try to obtain callers' personal information without their consent. But it means Yelp's call log tells you a call happened and nothing that lets you find the caller in your customer file. Two practical consequences follow.

• Export or copy the call report regularly. If call reporting is ever switched off, Yelp deletes the history. A quarter of call data you meant to reconcile later is gone.
• Get the full number from your own records, with the caller's knowledge. The simplest way is to confirm a callback number on every call, which most callers expect when they are asking for a quote. Your phone system or VoIP provider also logs inbound calls with a time and, usually, the caller's number; make one test call to your Yelp number from a mobile to see whether forwarded calls show the caller's own number. If they do, the call in your log at the same minute, with the same visible digits, is the Yelp call. Record the number with "Yelp" as its source.

The tempting shortcut — putting your own call-tracking number on the Yelp page — needs care. Yelp's guidelines for business information ask for a direct local number and say Yelp may not accept tracking numbers for third parties, and its own call reporting already replaces your number when switched on. Check before you swap anything. Our post on call tracking and revenue covers what a tracking number can and cannot tell you in general; for Yelp, the sanctioned route is Yelp's forwarding number plus your own call log.

And ask. A receptionist who asks every new caller how they found you, and writes the answer in the CRM, catches the Yelp callers who read your number off the page and dialed your main line from memory a week later. That is the same low-tech discipline that makes referral revenue visible, and it costs nothing.

Messages and quote requests: the email is Yelp's, not theirs

Request a Quote and messages arrive in your Yelp inbox. The customer's email address is replaced by a masked Yelp address that expires, and a phone number is included only if the customer opts in to phone or SMS contact. To match a quote request to a sale you need to record a real contact detail yourself.

This is the single most common way Yelp revenue goes missing in a reconciliation. A business copies quote requests into its CRM with the address Yelp shows, a string ending in messaging.yelp.com, and months later wonders why none of them matches an invoice. They cannot: the customer's invoice carries their real email. A Yelp relay address in a lead file is not a contact detail. It is a placeholder that looks like one.

1. When a quote request arrives with a phone number, record that number against the lead with "Yelp" as its source.
2. When it does not, ask for one in your first reply. A quote for a water heater or a roof needs a site visit or a call anyway, so the request is natural.
3. When the customer replies from their own email or calls you, add that real address or number to the same lead.
4. Never store the masked Yelp address as the customer's email. Keep it, if you like, in a notes field.

Whatever you collect, normalize the phone numbers before matching: Yelp, your phone system and your invoicing software will each write the same number differently, and formatting causes more missed matches than missing data does.

A Yelp quote request is matchable only once somebody at your business has written down a real phone number or email for it. The address Yelp shows you is a relay, and it will never match an invoice.

From Yelp's lead count to closed revenue

Cost per paying customer from Yelp is your Yelp Ads spend for a period divided by the number of customers whose first recorded contact came through Yelp and who then paid you. Getting there means discarding the leads that carry no contact detail, recovering the ones that can carry one, and matching those to sales.

Here is the plumbing company's quarter worked through. The figures are illustrative, built to show the shape of the calculation rather than taken from any account, but every step is one a real business can repeat with its own exports.

The arithmetic, step by step:

• Cost per Yelp-reported lead: $3,600 ÷ 420 = $8.57. This is the number on the dashboard, and it includes 240 leads (90 + 105 + 45) that no business could ever follow to a sale.
• Cost per lead with a real contact detail: $3,600 ÷ 144 = $25.00. Of 114 calls, 93 had their full number recovered from the phone log; of 66 quote requests and messages, 51 ended up with a real phone or email.
• Cost per paying customer: $3,600 ÷ 28 = $128.57. This is the figure to compare with every other channel.
• Close rates: calls 19 ÷ 93 = 20.4%, quote requests 9 ÷ 51 = 17.6%, together 28 ÷ 144 = 19.4%.
• Average sale: calls $24,700 ÷ 19 = $1,300; quote requests $13,950 ÷ 9 = $1,550.
• Revenue per dollar spent: $38,650 ÷ $3,600 = 10.7.

Two things stand out. The cost per paying customer is fifteen times the cost per reported lead, which is roughly what happens whenever a platform's definition of a lead is generous; we made the general case in cost per lead. And the 21 calls and 15 quote requests that never got a full contact detail are not zero. Some of them bought. Those sales landed in the unknown bucket, which is where they belong until somebody records the number — and they are an argument for recording it, not for estimating it.

Two rules keep the calculation honest. Match on the customer, not the order: a customer who first called you from a Google Business Profile listing in March and requested a Yelp quote in May was found by the profile, so pick one rule, such as the earliest recorded contact takes the credit, and apply it to every channel. Our post on duplicate leads explains why consistency matters more than which rule you pick. And leave the sales that match no lead where they are, in Direct / Unknown, rather than handing a share of them to Yelp because Yelp's dashboard is busy.

CloseRev reads exported files and has no connection to Yelp, so it sees Yelp only through the leads you record. Add each Yelp call and quote request to your lead file with the customer's real phone number or email — recovered from your phone log or CRM, never Yelp's relay address — and "Yelp" in the source column, and it matches them to closed sales on normalized phone and email, crediting each sale to the customer's earliest lead. A sale with no matching lead stays in Direct / Unknown.

Ads leads, organic leads and the listing you get anyway

Your Yelp page produces calls and quote requests whether or not you advertise. The honest question about Yelp Ads is what they add on top of the free listing, and the only way to see it is to compare periods with and without ads, measured in closed revenue rather than leads.

Yelp's split between ads leads and organic leads helps, but it does not settle the question. A person who clicked your ad might have found your free listing two results further down, and some of the organic leads may exist because the ad made your name familiar. The split describes the route; it cannot tell you what would have happened without the ad. In the worked quarter above, the $128.57 is a cost per Yelp customer, not a cost per customer the ads created, and the second figure is the one that decides the budget.

The test that answers it is simple and a little uncomfortable. Pause Yelp Ads for four to six weeks in a period that is not your peak, keep the listing as it is, and keep recording Yelp calls and quote requests exactly as before. Then compare the paying customers and revenue from Yelp in the paused weeks with the same weeks running ads, allowing for season. If Yelp revenue barely moves, the listing was doing the work. If it falls by more than the ad spend, the ads earn their keep. Our post on incrementality tests covers how to read a result like that without over-reading it.

Yelp's dashboard tells you how many leads clicked an ad first. Only a pause, measured in closed revenue, tells you how many customers the ads added to what the free listing would have brought.

What Yelp's numbers are still good for

The engagement figures that cannot be matched to a sale — map views, website clicks, bookmarks, check-ins — are still useful as a trend. A sudden fall tells you something changed on your page or in your category. They should be watched, never valued in dollars or added to a revenue report.

Some of the most valuable things Yelp does for a business are the ones nobody can attribute. A customer reads twenty reviews on Yelp, closes the app, searches your name on Google and calls the number from your Business Profile. That sale is real, Yelp helped, and no record anywhere connects the two. The same is true of a review that persuades somebody to call you rather than a competitor six months from now.

The temptation is to fill that gap with a guess: count Yelp page views as an assist, or give Yelp a percentage of unattributed revenue. Resist it. A guess that is written into a report looks exactly like a measurement and is much harder to remove than to add. Measure the Yelp revenue you can trace, month by month, read it beside your Yelp rating and review count over the same months, and be explicit that the halo is unmeasured. We set out why spreading the unknown across channels misleads in unattributed revenue.

The other place to compare is across pay-per-lead and listing channels. A Yelp customer at $128.57 is expensive or cheap only next to what a customer costs from Local Services Ads, Angi or Thumbtack — and the marketplaces have their own complication, because the request you paid for was often sold to competitors too, as we explain in lead marketplaces. Measured the same way against the same invoices, those comparisons are the only ones that should move budget. Businesses that rely heavily on Yelp, such as plumbers and other home-service trades, are the ones for whom this is worth doing every quarter.

The number to bring to your Yelp renewal

Bring cost per paying customer, by lead type, for the last two quarters, and the result of a pause test if you have one. Those are figures Yelp cannot compute because it never sees your invoices, and they turn a conversation about leads into one about customers.

Yelp account managers are measured on the numbers in the dashboard, and they will show you those numbers: leads up, cost per lead down, clicks on target. All of it may be true. None of it answers whether you should spend more. If you arrive with "Yelp calls close at 20% and cost us $129 a customer, against $210 from paid search", the conversation is about how to get more of what works. If you arrive with "I am not sure it is working", it is about the dashboard.

Yelp can tell you how many people looked, tapped and asked. Only your invoices can tell you how many paid, and the price per paying customer is the only Yelp number worth renewing on.