Where the data came from, and whether the accuracy claim means anything
Every data vendor publishes an accuracy percentage and almost none publishes the method behind it. The GTM Tech Index grades all 338 vendors on the same three questions: where the data came from and on what license, whether a performance claim carries a method, and what model produced the score. Here is what the market actually documents, with the vendors a buyer can check named by category.
Apollo.io, Bombora, GTM Workspace by ZoomInfo, Instantly, Leadfeeder, Saleshandy, Adapt.io, Clay, Clevenio, Clodura.AI, Cognism, DM Faster, Fuzzy Sequence, Hunter, Kaspr, Lead411, LeadPilot, Leadspicker, Lusha, Neuron, RB2B, Reply.io, Säljmotorn, Sendr, Signado, Surfe, Trigify, Vector, WaLead
Bombora, GTM Workspace by ZoomInfo, Leadfeeder, 6sense, Clevenio, Cognism, Lead411, Neuron, RB2B, Sendr, Signado, Surfe, Trigify, Vector, What's Trending
Clari Forecast, DealHub, Gong, GTM Workspace by ZoomInfo, HubSpot Sales Hub, Scratchpad, Adobe Marketo Engage, Airspeed, Alta, Demodesk, Endgame, Fullcast, GetAccept, Mediafly, Odie, Opine AI Presales Platform, Outdoo AI, Outreach, Playbook AI, Quantified AI, Revenue Grid, Revenue.io, Salesloft, SAP Sales Cloud, Veloxy
Across the whole index 29 of 338 vendors document sourcing, evidence and the model together and 145 document none of the three. The pattern is that evidence is strongest where the outcome arrives without the vendor's help. Revenue intelligence, where a quarter closes on its own schedule, documents its accuracy claims at 49 percent. Contact data, where accuracy is measured against a file only the vendor holds, documents its sourcing at 26 percent. What each of those 25 revenue intelligence platforms actually publishes is set out one by one below.
The three questions, index wide
Data Licensing and Provenance
Operational and Outcome Evidence
AI Disclosure and Model Transparency
Contact and enrichment data: who says where it came from
29 of 111 vendors that sell or resell B2B contact data document their sourcing and licensing well enough for a buyer to assess it. 54 document none of the three axes on this page.
- Apollo.io A
- Bombora A
- GTM Workspace by ZoomInfo A
- Instantly A
- Leadfeeder A
- Saleshandy A
- Adapt.io B
- Clay B
- Clevenio B
- Clodura.AI B
- Cognism B
- DM Faster B
- Fuzzy Sequence B
- Hunter B
- Kaspr B
- Lead411 B
- LeadPilot B
- Leadspicker B
- Lusha B
- Neuron B
- RB2B B
- Reply.io B
- Säljmotorn B
- Sendr B
- Signado B
- Surfe B
- Trigify B
- Vector B
- WaLead B
4 document all three axes on this page: Clay, Leadspicker, Reply.io, Säljmotorn.
Intent data: who says where the signal came from
15 of 43 intent providers document where their signals are collected. 7 document the model that turns a signal into a score.
- Bombora A
- GTM Workspace by ZoomInfo A
- Leadfeeder A
- 6sense B
- Clevenio B
- Cognism B
- Lead411 B
- Neuron B
- RB2B B
- Sendr B
- Signado B
- Surfe B
- Trigify B
- Vector B
- What's Trending B
No vendor in this category documents all three axes on this page. That is a finding about the category rather than about any vendor in it.
Revenue intelligence: who evidences the accuracy claim
25 of 51 revenue intelligence platforms publish a method behind their accuracy and outcome claims, and 19 document where the underlying data comes from.
A name answers who. This is what each one publishes, opened from the graded record. The full evidence, its source basis and the date it was verified sit on the vendor profile.
-
The first A on this axis in the index, and it is not close. Independent analyst evaluation across three firms and multiple categories: named a Leader in the first Gartner Magic Quadrant for Revenue Action Orchestration in December 2025 with Salesloft named a Visionary in the same grid; a Leader in Forrester's Revenue Orchestration Platforms evaluation in 2024; a Leader in Revenue Operations and Intelligence in 2022 with the highest ranked current offering among fourteen vendors evaluated; and a Strong Performer in Conversation Intelligence in 2023.
-
The deepest customer evidence in the index after Clari, and it is attributed rather than aggregated. Named enterprises appear with named individuals and their titles: a head of deal desk at Intuit, a global CPQ and CRM administrator at HP, a vice president of global deal desk at Braze, a global pricing product manager at The Adecco Group, a revenue operations manager at Zapier, and further named roles at Asure, Trintech and Verbit.
-
The evidence base here is the broadest graded in this index and it spans four independent kinds. Corporate performance is published and checkable rather than asserted: annual recurring revenue above 500 million dollars, growth above 55 percent year on year sustained across ten consecutive quarters, and roughly 584 million dollars raised.
-
The evidence base is the deepest in this index and most of it is verifiable outside the vendor's own control. The company is publicly listed, so its customer numbers, revenue and growth are filed and auditable rather than asserted.
-
The evidence base is deep and most of it sits outside the vendor's control. The company is publicly listed, so customer counts, revenue and retention are filed and independently auditable rather than asserted, and its scale claims of more than 299,000 customers across more than 135 countries are checkable against those filings.
-
The independent review base is the largest recorded in this index, and named executives attach figures to it. 8, with the vendor recognised there as both a leader and a high performer.
-
Interactive, an Australian IT services company, reports moving from 50 manually handled sales accepted leads to 200 to 400 per month after rebuilding Marketo, Microsoft Dynamics and its website together.
-
Better than most of this cohort, and short of the top band on basis. A named customer figure is published with a timeframe attached: an enterprise software customer reporting more than 193,000 dollars saved within 90 days and roughly six hours reclaimed per sales representative per week.
-
5x more email replies stated as measured across 2M+ prospects, 72% faster lead response, sub-30-second inbound response, 20 hours saved per rep per week, one customer reaching seven-figure pipeline in six months with a one-person GTM team.
-
Breadth and attribution are both strong. More than a dozen case studies are published and linked individually, each attributed to a named person with their title and employer, among them a vice president of revenue at OMR Reviews, a business development lead at StudySmarter, a vice president of sales at WorkFlex, a vice president of strategy at HeyJobs, a head of sales at Tanso and a head of inside sales at Treatwell, alongside a logo set including Verivox, Forto, Holidu, Pipedrive, AUTO1 and WorkMotion.
-
Three named customers, each attributed to a named individual with title and employer, each with its own linked case study: a revenue vice president at Hex, a co founder and chief operating officer at Monte Carlo, and a revenue enablement lead at Retool.
-
The named evidence here is among the strongest in the index. Fourteen enterprise logos are published including Gong, Rubrik, Lenovo, Thermo Fisher, Asurion, TechTarget, Urban Outfitters and a global consultancy, three case studies carry a named person with their title and employer, and eight testimonials each carry a name, a role and a company rather than a first name and an initial.
-
The distinguishing strength here is that the names and the numbers meet, which is exactly what a vendor graded earlier in this same session failed to do despite a larger logo wall.
-
Sealed Air consolidated four libraries and reports 90 percent user adoption within two months, with engagement per user up 13 percent. In 2020, it selected Mediafly over Showpad while introducing value calculators.
-
7 percent increase in output quality, from benchmarking complex queries against 70 production customer instances, with output quality combining usefulness, grounding, directness and specificity.
-
Among the better evidenced young vendors in this index. Four case studies are published on named and recognisable customers, each carrying a quantified claim in its own title, covering three times more closed won revenue per sales engineer hour at one, over a hundred hours saved at another, and proof of concept volume up eighty five percent at a third.
-
Among the strongest evidence bases in this index and short of the top band only on method. Three testimonials carry a full name, a role and a named recognisable employer, each attached to a specific quantified result, covering better objection handling at a listed insurer, reduced call administration at a major infrastructure company, and a single coaching framework across three regions at an events platform.
-
The customer roster is the strongest in this index and the measurement behind the claims is not published. Twelve named enterprises appear on the enterprise page alone, spanning vehicle rental, data infrastructure, scientific publishing, conferencing, industrial manufacturing, cloud data warehousing, healthcare distribution, identity management, business software, data resilience, network management and information technology service management.
-
A genuinely substantial evidence base for a vendor of this size, with real defects in it. Three case studies each name a company, name an individual with their title, carry a specific figure and link to a full write up, covering fifty two percent more meetings booked per representative at an outsourcing firm, twenty one percent more revenue per representative at a consultancy, and complete follow up of marketing qualified leads at a sales platform.
-
Seventeen customer logos are named and several are among the largest pharmaceutical companies in the world, and seven testimonials carry a full name, a specific job title and a named employer, with headshots, across four different customers.
-
Thirteen named case studies are organized by industry in the navigation itself, a national law firm with caseload up fifteen to twenty percent, a consulting major that rebuilt its sales model, a capital markets firm, a training company, a software company, and named individuals with roles carry quantified quotes, thirty to forty percent of reporting time saved at a named consultancy.
-
The reference set is enterprise grade and specific: a global hardware company selling into one hundred fifty plus countries reports four hundred percent more opportunities, a logistics firm reports eight times more conversations and ten times more meetings, an equipment rental major reports every call coached and scored, and named individuals with titles from a payments company and a real estate technology firm carry quoted endorsements, one stating the platform was chosen twice over a named competitor.
-
The corroboration is broad and the headline numbers are the weakest part of it. What is strong: nine named customers displayed on the pricing page, including a payments company, a global technology manufacturer, a recruitment platform, a delivery marketplace, an industrial conglomerate and a commerce platform.
-
The independent review base is the largest recorded in this index, and the vendor's own outcome claims are the thinnest part of the record. 4 percent of reviewers, which lets a buyer weight the ratings against their own size rather than reading an undifferentiated average.
-
A strong review base carries this, and the headline claims do not. The reviews are substantial and unusually distributed: four point seven out of five across two hundred and thirty four entries on the principal enterprise platform, with sixty percent of reviewers from enterprise organisations.
7 document all three axes on this page: Adobe Marketo Engage, Demodesk, Gong, Mediafly, Revenue.io, SAP Sales Cloud, Scratchpad.
What to ask when the sourcing is not published
Each carries a different obligation for the buyer, and a provider that will not name the mix is asking you to accept all of them.
Contribution pricing is often the real price. The discount is paid in your own contact records, and that is a decision your privacy counsel should make rather than your procurement team.
Continuous and real time are not measurable claims. A monthly re verification share is, and a provider that tracks it can state it.
An accuracy rate on a vendor chosen sample tells you about the sample. Ask for the denominator before you ask for the percentage.
68 of 338 vendors disclose the model behind a score. A score whose inputs you cannot name cannot be diagnosed when it stops working.
This is the question the category does not volunteer. Beating a sales leader's submitted number is a low bar; beating last quarter plus growth is the one that justifies the license. 25 of 51 revenue intelligence platforms publish any measurement basis at all, so for the majority there is nothing to compare against either baseline.
This costs the buyer an afternoon and settles more than a procurement cycle of reference calls. Reluctance is itself an answer.
Common questions
Where do B2B contact data providers source their data and is it licensed?
Most do not say. Of the 111 vendors in the GTM Tech Index that sell or resell B2B contact data, 29 document their sourcing and licensing at a standard a buyer can check, 26 percent, and 15 carry a D, meaning the question is not addressed at all. The 6 graded at the top of the scale are Apollo.io, Bombora, GTM Workspace by ZoomInfo, Instantly, Leadfeeder, Saleshandy. Where sourcing is disclosed, the common structure is a mix of public web collection, user contributed data from a community or extension, licensed third party files, and partner contribution networks where customers exchange their own contact records for access to everyone else's. Those four carry very different obligations for the buyer, which is why the GTM Tech Index grades the disclosure rather than the sourcing method: a provider that names its inputs can be assessed, and one that does not is transferring that assessment to you.
How do you evaluate the accuracy of a B2B contact data provider?
Never on the headline percentage, because almost every provider publishes one and 80 of 111 publish it with no method attached. Of the 111 contact data providers in the GTM Tech Index, 31 state a measurement basis a buyer could reproduce, 28 percent, and 23 carry a D. Three things make a claim assessable. First, the denominator: accuracy measured against records the provider chose to sample is not accuracy across the file you will actually query. Second, the date and the method of verification, since a bounce tested address and a pattern generated one can both be called verified. Third, the segment, because coverage in North American mid market technology tells you nothing about European manufacturing. The practical test costs nothing: send the provider a list of 200 accounts you already know well, ask for their records against it, and measure the result yourself. A provider confident in its file will run that test. Treat reluctance as the answer.
Which B2B contact data providers publish the basis for their accuracy claims?
31 of the 111 vendors in the GTM Tech Index that sell or resell B2B contact data publish a basis for their accuracy and outcome claims at a standard a buyer can check, 28 percent, and 23 carry a D, meaning a claim with nothing published behind it. 5 carry an A from the GTM Tech Index, the claim published with its measurement basis: GTM Workspace by ZoomInfo, Saleshandy, Smartlead, Snov.io, Unify. A further 26 publish a basis that is checkable without being fully reproducible, among them Amplemarket, Apollo.io, Autobound, Clay, Cognism, Coldreach, Crono, Dripify, Evergrowth, FirstTouch, and 16 more. Everyone else publishes a headline percentage with no stated method, or no figure at all. Four things turn an accuracy number into a claim a buyer can test: what it was measured against, on how large a sample, on what date, and in which segments. A provider that states all four has given you something to check in a trial. One that states none has given you a slogan.
How often do B2B contact data providers refresh their records?
A refresh cadence is a claim like any other, and the GTM Tech Index grades whether the claim carries a method rather than whether the number is impressive. Continuous, real time and always on are the phrases used most often across the 111 vendors in this lane, and they are not verifiable statements. 31 of them publish an evidence standard behind their operational claims, so the great majority of published refresh claims have nothing behind them. The questions that turn the claim into something checkable are these: what triggers a record to be re verified, is it a fixed calendar cycle or an observed signal such as a job change; what share of the file is re verified in a given month rather than merely re scored; how long a departed contact stays live before removal; and whether the same cadence applies to email addresses, phone numbers and firmographics, which usually decay at very different speeds. Ask for the monthly re verification share as a number. Vendors that measure it will tell you, and vendors that do not will describe the process instead.
How do B2B intent data providers differ in how they collect intent signals?
They differ more than the category's marketing suggests, and the difference is mostly undisclosed. Of the 43 intent providers in the GTM Tech Index, 15 document where their signals come from at a checkable standard, 35 percent, and 4 carry a D. The ones that do document it are Bombora, GTM Workspace by ZoomInfo, Leadfeeder, 6sense, Clevenio, Cognism, Lead411, Neuron, RB2B, Sendr, Signado, Surfe, Trigify, Vector, What's Trending. The collection methods behind the category are genuinely different products: bidstream data observed through advertising exchanges, publisher cooperatives where a network of trade sites shares reader behavior, first party signals from your own site and product, review site activity, and inferred signals derived from hiring, funding or technology change. Bidstream and cooperative data describe an anonymous company; first party data describes a known person. The model that turns any of these into a score is disclosed by 7 of 43. Not one vendor in this lane documents all three of provenance, evidence and the model behind the score. Ask which of those five inputs is in the feed you are buying, in what proportion, and what happens to the score when one goes away.
Which B2B intent data providers publish how they measure accuracy?
11 of the 43 intent and signal providers in the GTM Tech Index publish a measurement basis behind their accuracy and outcome claims at a standard a buyer can check, 26 percent, and 7 carry a D, a figure with nothing behind it. 3 carry an A from the GTM Tech Index, the claim published with its method: GTM Workspace by ZoomInfo, Nooks, Unify. A further 8 publish a basis that is checkable without being fully reproducible: Autobound, Clickback, Cognism, Coldreach, Crono, MarketBetter, Selling.com, Signado. The rest cite accuracy without saying what it was measured against. That matters more for intent than for contact data, because an intent score has no ground truth a buyer can look up: a contact record is right or wrong, but a signal is only right if the account goes on to buy. The only honest test is a comparison of outcomes, whether the accounts the model flagged converted at a higher rate than the ones it did not, over a stated period and sample. Ask for that comparison, then run a blind test on accounts whose buying status you already know.
Which revenue intelligence platforms have evidence their forecasts are accurate?
The GTM Tech Index grades all 51 revenue intelligence platforms on this question, and 25 of them publish evidence behind their accuracy and outcome claims at a standard a buyer can check, 49 percent, against 28 percent for contact data providers and 26 percent for intent providers. 6 carry an A from the GTM Tech Index, the top of the scale: Clari Forecast, DealHub, Gong, GTM Workspace by ZoomInfo, HubSpot Sales Hub, Scratchpad. A further 19 publish a basis that is checkable without being fully reproducible: Adobe Marketo Engage, Airspeed, Alta, Demodesk, Endgame, Fullcast, GetAccept, Mediafly, Odie, Opine AI Presales Platform, Outdoo AI, Outreach, Playbook AI, Quantified AI, Revenue Grid, Revenue.io, Salesloft, SAP Sales Cloud, Veloxy. Everything outside those 25 publishes an accuracy or outcome figure with no method attached, or publishes nothing at all. That is a materially better record than contact data or intent, and the likely reason is that forecast accuracy is measured against an outcome that arrives on its own, the quarter closes whether or not the vendor wants it to, whereas contact accuracy is measured against a file only the vendor holds. Two cautions before treating a published figure as settled. Forecast accuracy is usually reported as a variance against the vendor's own submitted forecast rather than against the naive baseline of last quarter plus growth, and a model that beats a rep's gut but not the naive baseline has not earned its license. And 13 of 51 disclose what the model is, so the number is often published without the method that produced it. 7 platforms document provenance, evidence and the model together: Adobe Marketo Engage, Demodesk, Gong, Mediafly, Revenue.io, SAP Sales Cloud, Scratchpad. All 25 are listed below with what each one actually publishes, because a shortlist of names is only half of this question: the useful half is the evidence sitting behind each name, and whether it is an analyst evaluation, an audited financial filing, a named customer with a figure attached, or a review base large enough to mean something.
How does the GTM Tech Index grade data provenance and evidence?
On three axes, applied identically to all 338 vendors: Data Licensing and Provenance for where the data came from and on what license, Operational and Outcome Evidence for whether a performance or accuracy claim carries a method, and AI Disclosure and Model Transparency for what produced a score. A grade records what the vendor has published, so a low grade means the public record is thin rather than that the practice is bad, and a vendor may well answer in a contract what it does not answer on its site. Across the index, 29 of 338 vendors document all three and 145 document none. AI Disclosure and Model Transparency is the thinnest of the three at 20 percent with 57 vendors at a D or below. Grades are never aggregated into a composite score, and every grade carries a source basis and a verification date on the vendor profile.
How to read these numbers
Documented means a grade of A or B: substantive disclosure a buyer can check, with minor gaps at worst. C means claims without documentation. D means a material claim with no verification path, or silence where the product plainly requires an answer. A low grade records a thin public record rather than a bad practice, and a vendor may answer in a contract what it does not answer on its site. A grade belongs to a vendor rather than to one of its features, which matters here because many of these vendors appear in more than one category and the grade may have been earned on a different product line. Category populations on this page count primary and secondary membership together, because selling contact data is a role a vendor can fill without it being their main business; the same category on the capability axes page is primary only and will show a smaller roster. Figures generated from the live index on 2026-10-14 across 1014 grades. See the Methodology for verification standards.
Work from the data
Best B2B data and enrichment vendors: the shortlist, decided on the axes that matter in this category.
The outreach compliance framework: whether you are permitted to use the data, which is a different question from whether it is any good.
What to look for in go to market software: all seventeen axes, with the share of each category that documents each one.