Statistics
Yelp statistics for local business, traced to source
Real numbers from Yelp's own SEC filings, investor disclosures and official Trust & Safety Report — the single most directly verifiable source type used anywhere on this site.
Most statistics on this site trace to surveys, platform-usage data, or expert opinion. This page leans differently: as a publicly traded company, Yelp is legally required to disclose real operational and financial metrics every quarter.
It separately publishes an annual Trust & Safety Report and a quarterly Consumer Alerts page that name real enforcement actions rather than aggregated percentages. Together, that makes this some of the most directly verifiable data available anywhere in local marketing research — and it also makes clear, in Yelp's own words, exactly what the company chooses not to disclose. The grading scale used here is set out in our research methodology.
The most useful finding across the whole page isn't any single figure. It's the pattern: Yelp discloses precise absolute counts — advertiser locations, accounts closed, reviews filtered for a specific reason, countries where operations exist — and almost never a headline rate or percentage.
That's worth knowing before quoting any 'X% of Yelp reviews are fake' statistic you find elsewhere, because Yelp itself has never published one. For what reviews across all platforms actually do to a customer's decision, the local review statistics page is the better starting point.
How the grading works
The grading is deliberately unforgiving. A statistic from a good organisation still earns a C if the sample size is not published beside the number, because a reader cannot check what they cannot see.
Cite freely
Primary source, dated, with sample and method published. Quote it, with its scope attached.
Cite with scope
Real research with a published sample, but narrower than the way it is usually quoted — self-reported, or one country, or one kind of business.
Cite with caution
Named source and plausible method, but the sample size or methodology is not published where the number is. Attribute it, do not lean on it.
Do not cite
Untraceable, undated, or superseded by later changes to the thing it measured. These are the numbers keeping the roundups alive.
How many businesses actually pay to advertise on Yelp — from Yelp's own regulatory filing
As a public company, Yelp discloses real operational numbers every quarter to the SEC — a genuinely different, more verifiable source type than almost anything else cited on this site. For how Yelp sits among the directories a local business is actually listed on, see the local citation statistics.
510,000
Grade APaying Advertising Locations on Yelp, as of Q2 2026.
What it does not say A public company's SEC filing is about as verifiable as a source gets — legally required accuracy, audited financials, and genuine consequences for misstatement. This is a real count of paying advertisers, not a survey estimate. Yelp's own investor fast-facts page independently repeats the same figure for the same quarter.
-3%
Grade Ayear-over-year change in Yelp's advertising revenue in Q2 2026, even as overall net revenue grew 1%, driven by a 98% jump in other revenue streams (data licensing, food ordering and similar).
What it does not say Worth noting for context: Yelp's core local-business advertising revenue is currently shrinking even as the company overall grows, which may reflect broader shifts in where local businesses are choosing to spend on visibility.
$91.4 million
Grade Aadjusted EBITDA in Q2 2026, a 24% margin — down 9% year over year even as net revenue grew.
What it does not say Read against the advertising-revenue decline above, this is the same story from the profitability side: the business is growing on paper while its most profitable, longest-standing revenue line contracts.
5,168
Grade Aemployees globally, as of 31 December 2025 (including employees on leave).
What it does not say Yelp does not break this figure down by country in the filing, so it can't be used on its own to size the US share of the workforce — but read alongside the operating-footprint finding further down this page, it fits the same pattern of a company that is smaller and more US-concentrated than its consumer-facing brand recognition suggests.
What this means for a business deciding whether to advertise
The two headline numbers above point in different directions, and both matter to the same decision. A 510,000-location advertiser base is real scale — Yelp is not a niche product for local businesses to consider.
But advertising revenue specifically fell 3% year over year while adjusted EBITDA margin compressed, in the same quarter Yelp's total business grew on the back of licensing and food-ordering revenue rather than local advertising.
That combination — a large, established advertiser base whose spend is now shrinking as a share of a growing company — is the single most useful piece of context on this page for anyone weighing Yelp against other places to spend a local-marketing budget.
It doesn't say whether Yelp works for a specific business or category; it says the platform-wide trend a salesperson is unlikely to volunteer is running against the product they're selling. Where a directory earns its keep for most local businesses is as a consistent listing rather than as an ad buy — the case for that is set out in the local citation building guide.
How many reviews Yelp actually has, and how fast that number is growing
Cumulative review counts get quoted constantly and rarely dated. Yelp's own investor materials and its 2025 Trust & Safety Report give two figures a year apart that agree with each other almost to the review — which is itself a useful cross-check on both documents.
330 million
Grade Acumulative reviews contributed to Yelp, as of 31 December 2025 — including reviews currently recommended, not recommended, or since removed.
What it does not say Yelp's own definition is worth repeating exactly: this is every review ever contributed, not the number currently live and recommended on the site. Treating 330 million as 'reviews you can currently read' overstates what a visitor actually sees.
22 million
Grade Areviews were contributed to Yelp during 2025 alone.
What it does not say This figure is not just plausible against the 330 million cumulative total above — it is arithmetically consistent with it. Yelp's own Wikipedia-cited 10-K disclosure put cumulative reviews at roughly 308 million as of 31 December 2024; 308 million plus 22 million lands almost exactly on the 330 million figure disclosed for the end of 2025, from an entirely separate Yelp document. Two independently published company figures agreeing to within rounding is a stronger form of verification than either alone.
28 million
Grade Bmonthly average app unique devices in 2025.
What it does not say This is the current form of Yelp's traffic disclosure — a mobile-app device count, not a combined web-and-app 'unique visitors' figure of the kind Yelp reported in earlier annual filings. Comparing this number against an older 'unique visitors' statistic from a pre-2025 source compares two different metrics, not the same metric over time.
446 characters
Grade Baverage length of a Yelp review currently classified as recommended.
What it does not say Yelp doesn't publish the equivalent average for reviews that are not recommended, so this figure can describe the reviews the recommendation software favours, but not whether length itself is a factor it weighs — those are different claims, and only the first is supported here.
What this means for how much review volume actually matters
330 million reviews sounds like saturation, but the more useful number on this page is the growth rate: 22 million added in a single year against a base that took two decades to build.
For an individual local business, the practical takeaway isn't the platform total — it's that the review count on a competitor's profile is not a fixed target. A gap that looks unbridgeable today is being closed, business by business, at a pace roughly a fifteenth of the entire platform's history every twelve months. Doing that deliberately and within the platforms' own rules is the subject of the local review strategy guide.
The 28 million monthly app devices figure is the more relevant number for a specific decision — how many people are actually opening the app in a given month, category and city. That is not something this page, or any platform-wide figure, can answer for you: it needs a category-specific check inside Yelp for Business, not a blended national average.
Yelp's review filter, and the trust-and-safety numbers it does — and doesn't — publish
This is the most-asked question about Yelp and the least precisely answered by Yelp itself. Its own documentation explains the filter's logic in detail and, deliberately, never states what share of reviews it catches. What it does publish, in an annual Trust & Safety Report, is a set of absolute enforcement counts — a different and in some ways more useful kind of transparency.
No published rate
Grade Ais disclosed by Yelp for the share of reviews its recommendation software filters out — despite being one of the most frequently quoted 'facts' about the platform.
What it does not say Yelp's own page states that the software runs on every review, uses 'hundreds of signals of quality, reliability, and user activity', and that 'no employee at Yelp has the ability to override the decisions that the software makes' — but it stops short of a headline percentage. A 2013 New York Attorney General press release announcing a settlement with Yelp and 18 other companies over fake reviews quotes Yelp's own senior litigation counsel describing the filtering approach in similar terms, and likewise states no percentage.
~500,000
Grade Asuspected AI-generated reviews Yelp says it filtered out during 2025.
What it does not say Against the 22 million reviews contributed in 2025 (see the section above), this is roughly one AI-suspected review filtered for every 44 submitted — a ratio this page has calculated from two of Yelp's own disclosed figures, not one Yelp states directly, and it covers only the AI-generated category specifically, not the total filtered-review rate.
1.3 million+
Grade Auser accounts Yelp says it closed during 2025 — a 138% increase on 2024 — including 889,800 fake phone-support accounts tied to a coordinated airline customer-service scam, and 1,991 accounts closed specifically for participating in review-exchange rings.
What it does not say The scale of the year-over-year jump is worth reading carefully rather than as straightforward evidence platform abuse doubled: a large single-category enforcement action, like the airline-scam accounts named here, can move a total like this on its own without reflecting a broader change in ordinary review behaviour.
50,700+
Grade Apotential new business pages Yelp says it rejected during 2025, a 29% increase on 2024, alongside 46,200+ reviews removed for containing inappropriate content and 3.6 million+ photos removed, a 25% increase on 2024.
94
Grade Abusinesses carried a Suspicious Review Activity Alert and 21 carried a Compensated Activity Alert in the most recent quarterly cycle published, against a cumulative 6,300+ such alerts placed since the programme began in 2012.
What it does not say This is one quarter's worth of new alerts, not an annual or cumulative figure — roughly 115 new alerts in the cycle. Multiplied across four quarters, that lands close to the 491 Compensated and Suspicious Review Activity Alerts the 2025 annual report states for the full year, which is a reasonable internal consistency check between Yelp's own quarterly and annual disclosures.
| Action | 2025 count | Change on 2024 |
|---|---|---|
| Suspected AI-generated reviews filtered | ~500,000 | Not disclosed |
| User accounts closed | 1,300,000+ | +138% |
| Photos removed | 3,600,000+ | +25% |
| Reviews removed for inappropriate content | 46,200+ | Not disclosed |
| New business pages rejected | 50,700+ | +29% |
| Reports made to other platforms | 1,020+ | +45% |
| Media Attention Alerts placed | 1,450+ | Not disclosed |
| Compensated / Suspicious Review Activity Alerts placed | 491 | Not disclosed |
All figures as stated by Yelp for calendar year 2025; year-over-year change shown only where Yelp itself published a comparison to 2024. None of these figures is a percentage of total review volume — they are absolute enforcement counts. Source: Yelp Inc., 2025 Trust & Safety Report
What this means for the "X% of reviews are fake" claims you'll see elsewhere
The absence of a published filter rate is not an oversight. Yelp's own documentation, and its response to a 2013 New York Attorney General investigation, both describe the filtering approach in detail while stopping short of a headline percentage in either case.
That means every "18% of Yelp reviews are filtered" or "1 in 4 reviews doesn't make the cut" figure circulating online is not coming from Yelp. It traces, where it traces anywhere at all, to independent research that sampled a specific city over a specific window years ago, not to an official, current, platform-wide rate.
The absolute 2025 enforcement numbers above are a genuinely useful substitute: they show the scale and direction of Yelp's own moderation activity — up sharply across most categories in 2025 — without the false precision of a percentage nobody at Yelp has actually published.
Treat the counts as evidence of enforcement intensity, not as a base for computing what share of any individual business's reviews might be affected. What is measurable is how reviews affect the choice itself, which the local review statistics page covers across platforms — the TRUSTED stage of the FOUND → TRUSTED → CHOSEN framework.
Is Yelp a global platform, or a US product with an app available elsewhere?
Yelp's own marketing lists a long roster of countries. Its SEC filings describe something considerably narrower. Both are true simultaneously, and the gap between them is the more useful fact for anyone outside the US weighing whether Yelp is worth their time.
32 countries
Grade Aincluding the United States, are listed on Yelp's own investor fast-facts page as places the platform is available — Argentina, Australia, Austria, Belgium, Brazil, Canada, Chile, Czech Republic, Denmark, Finland, France, Germany, Hong Kong, Italy, Japan, Malaysia, Mexico, New Zealand, Norway, Philippines, Poland, Portugal, Republic of Ireland, Singapore, Spain, Sweden, Switzerland, Taiwan, the Netherlands, Turkey, the United Kingdom and the United States.
What it does not say India is not on this list. Nor is China, most of Africa, or most of South Asia and the Gulf. For a site whose own research also covers India's local-business market, that's a directly relevant absence rather than an incidental one.
5 countries
Grade AUnited States, United Kingdom, Canada, Ireland and Germany — are the only ones Yelp's own Q2 2026 SEC filing describes when stating where the company 'has operations'.
What it does not say This is a narrower, legally scoped statement than the 32-country availability list above — 'has operations' in an SEC filing typically means a substantive business presence (offices, employees or revenue reporting), not merely that the consumer product is technically accessible from that country.
What this means for a business outside the United States
The two figures above describe the same company from two different legal and marketing registers, and reading only one of them gives a misleading picture.
"Available in 32 countries" is true and is also compatible with a product that has, in SEC-filing terms, meaningful operations in five. That pattern is common to many US-headquartered consumer platforms, where the app or website works globally on the open internet long before the company builds local sales, support or moderation infrastructure in a given market.
For a business in a country outside that five — including India, which appears in neither list — the practical read is that Yelp is not currently a platform with a dedicated local operation behind it, whatever the app store listing suggests.
That's a fact about company footprint, not a verdict on consumer usage in any specific country. No data in Yelp's own disclosures, or identified in the sources reviewed on 8 September 2026, states how many of the app's 28 million monthly unique devices are outside its five operating countries.
For a market-specific view of how local discovery actually works in India, see this site's own India local business statistics page, which covers the platforms India's local searches actually run through. The wider local search behaviour statistics cover where people start looking in the first place.
What the evidence covers
Every figure on this page describes Yelp's own global, company-wide disclosures — not a country-specific breakdown, because Yelp's public filings mostly don't provide one.
Where the evidence points specifically to the United States, the United Kingdom, Canada, Ireland or Germany — the five countries Yelp's own SEC filing names as places it 'has operations' — that is stated explicitly in the relevant record above.
India is not among Yelp's disclosed operating countries, nor among the 32 countries its investor materials list as places the platform is available. No substantial India-specific Yelp usage data was identified in the sources reviewed on 8 September 2026.
The numbers that did not survive the trace
These appear in current statistics roundups. Each one fails on provenance, and quoting them transfers the failure to you.
~20–25%
Do not citeof all Yelp reviews are filtered out by the recommendation software, in any given year.
Traced to No single origin — the figure varies by source, and different articles attach different numbers (commonly cited ranges include 16%, 18%, and 'roughly a quarter') to the same underlying claim, none linking back to a current, platform-wide Yelp disclosure.
Yelp's own documentation on how the recommendation software works, and its 2013 statement to the New York Attorney General on the same subject, both describe the mechanism without stating a percentage. Where a specific number is attached at all, it typically traces to independent academic sampling of one city over a fixed historical window — a real measurement of a real sample, but not a platform-wide, current rate, and not Yelp's own figure. Quoting it as 'what percentage of Yelp reviews get filtered today' overstates what any traceable source actually supports.
5–9%
Do not citeincrease or decrease in a restaurant's revenue for every one-star change in its Yelp rating.
Traced to Widely attributed to Harvard Business School research on Yelp ratings and independent-restaurant revenue, circa 2011–2016.
This page could not independently verify the figure against a primary source during the research pass completed on 8 September 2026 — the university and publisher pages hosting the original paper returned access errors rather than the study itself. The claim may well be accurate; it has circulated in reputable outlets for over a decade. Until it can be traced to and read directly from the primary paper rather than a secondary description of it, it doesn't meet this site's citation standard, and it is listed here rather than presented as a graded, cite-ready statistic.
$53
Do not citeaverage cost per lead for Yelp Ads, or similarly specific dollar figures for Yelp advertising ROI, appearing in various marketing blog posts.
Traced to Not traceable to any Yelp disclosure or independent, sampled study identified in the sources reviewed on 8 September 2026.
Yelp does not publish an average cost-per-click, cost-per-lead or ROI figure for its advertising product in its investor materials — the SEC filings on this page disclose Paying Advertising Locations and revenue, never a per-lead cost. Any specific dollar figure for Yelp Ads' cost per lead currently circulating is not attributable to Yelp itself or to a published, sampled study of Yelp advertisers specifically.
How this page is made and how to cite it
Figures on this page come from four kinds of Yelp self-disclosure, graded accordingly: its SEC filings (Form 10-Q and 10-K, the highest-verifiability source type used anywhere on this site), its investor fast-facts page, its annual Trust & Safety Report, and its quarterly Consumer Alerts page.
Two calculated figures are marked explicitly as this page's own arithmetic on Yelp's disclosed numbers, not as numbers Yelp itself states.
Last verified: 8 September 2026. Reviewed quarterly, matching Yelp's own reporting cadence, and re-checked in full whenever Yelp publishes a new Trust & Safety Report.
Preferred citation Parashar, S. (2026). Yelp statistics for local business, traced to source. shrutiparashar.com/statistics/yelp-statistics/
Questions
- How many businesses actually pay to advertise on Yelp?
- 510,000 Paying Advertising Locations as of Q2 2026, per Yelp's own quarterly SEC filing and independently repeated on its investor fast-facts page — a directly verifiable regulatory disclosure rather than a survey estimate.
- Is Yelp's advertising business growing or shrinking?
- By Yelp's own Q2 2026 filing, advertising revenue specifically declined 3% year over year, even as the company's total net revenue grew 1%, driven by a 98% increase in other revenue streams like data licensing and food ordering. Adjusted EBITDA fell 9% year over year in the same quarter. Yelp's core local-advertising business is currently contracting even as the company overall grows elsewhere.
- How many reviews does Yelp have in total?
- 330 million cumulative reviews as of 31 December 2025, according to Yelp's own investor fast-facts page — a figure that includes every review ever contributed, whether currently recommended, not recommended, or since removed, not the number of reviews a visitor can currently read. Yelp's 2025 Trust & Safety Report separately states 22 million reviews were added during 2025 alone, which is consistent with the year-over-year change in the cumulative total.
- What percentage of Yelp reviews get filtered out by the recommendation software?
- Yelp does not publish this figure. Its own official documentation on how the recommendation software works, and its 2013 statement to the New York Attorney General on the same subject, both describe the filtering logic without stating a percentage. Figures like '18%' or '1 in 4' circulating in other roundups are not Yelp's own disclosure — where traceable at all, they come from independent academic sampling of a single city years ago, not a current platform-wide rate. What Yelp does publish is absolute enforcement counts: roughly 500,000 suspected AI-generated reviews filtered in 2025, for example.
- Does Yelp disclose how many people use its app?
- Yes, as a monthly average: 28 million app unique devices in 2025, per Yelp's investor fast-facts page. This is narrower than the combined web-and-app 'unique visitors' figures Yelp reported in some earlier annual filings, so a current app-devices figure and an older unique-visitors figure from a different year are not directly comparable metrics.
- Is Yelp available in India?
- India does not appear on Yelp's own investor fast-facts page, which lists 32 countries where the platform is available — the United States, Canada, most of western Europe, and a scattering of markets across Latin America and Asia-Pacific, but not India, China, or most of South Asia. No data on an India-specific user base or business listing count was identified in Yelp's disclosures reviewed on 8 September 2026.
- Is Yelp actually a global company, or mostly a US business?
- Both descriptions are true at once, depending on which of Yelp's own documents is doing the describing. Its investor fast-facts page lists 32 countries where the platform is available. Its Q2 2026 SEC filing, a more legally scoped document, states that Yelp 'has operations' in only five: the United States, United Kingdom, Canada, Ireland and Germany. Being technically accessible in a country and having a substantive business presence there are different claims, and Yelp's own filings draw that line explicitly.
- How does Yelp's recommendation software decide which reviews to show?
- According to Yelp's own Trust & Safety documentation, the software runs continuously on every review submitted and scores it against 'hundreds of signals of quality, reliability, and user activity', looking specifically for conflicts of interest, solicited reviews, low community-reliability signals and low-usefulness content. Yelp states that no employee can override the software's decisions, and that it applies uniformly regardless of whether a business advertises. A review moved to 'not recommended' status stays accessible via a separate page but doesn't count toward the business's star rating or review count, and its status can change later as the software gathers more signal.
- Has Yelp published data on AI-generated fake reviews specifically?
- Yes. Yelp's 2025 Trust & Safety Report states it filtered out approximately 500,000 suspected AI-generated reviews during 2025 — a category the report treats separately from its other enforcement figures, such as the 46,200+ reviews removed for inappropriate content or the 1.3 million-plus user accounts closed that year. Against the 22 million reviews contributed in 2025, that AI-suspected figure works out to roughly one in every 44 submissions, though Yelp itself doesn't publish that ratio directly — it's a calculation this page has made from two of Yelp's own separately disclosed figures.
- Is the claim that a one-star Yelp rating change moves restaurant revenue by 5–9% real?
- It's a widely repeated figure attributed to Harvard Business School research on Yelp ratings, but this page could not independently verify it against the primary paper during the research pass completed on 8 September 2026 — the hosting pages returned access errors rather than the study itself. It is listed on this page as a claim under review rather than presented as a graded, cite-ready statistic, in line with this site's rule against citing a study it hasn't been able to read directly.
- Where does Yelp's revenue actually come from, if not mostly advertising growth?
- Yelp's Q2 2026 net revenue of roughly $376 million grew 1% year over year, but that growth was driven by a 98% jump in 'other revenue' — the category covering data licensing, food ordering integrations and similar non-advertising products — while advertising revenue itself fell 3% over the same period. Yelp does not break out other revenue or advertising revenue by exact dollar figure in its press release, only the percentage changes, so the precise split between the two lines can't be stated more exactly than that from the sources reviewed here.
Related reading: how provenance is judged here , local review statistics , India local business statistics .
This is one page of a graded statistics library that also names the 68 widely repeated claims it could not trace.
