B2B intent data promises to tell you which accounts are ready to buy. Many teams pay for it and still cannot say what it changed.
ARS B2B Social Bridge read the sources behind the claims and wrote it up as a free 16-page PDF, the B2B Intent Data Reality Check 2026. This article is the short version. Here’s the thing: almost every accuracy number we found comes from a company that sells something near intent data, so we badge each one. It is general information, not legal or investment advice.
Get the full B2B Intent Data Reality Check
Sourced numbers, a bias map, a signal-to-action ladder, ten vendor questions and a 30-day pilot. Every figure is badged by who published it.
- Bias map
- Source badges
- Privacy
- Action ladder
- Vendor questions
- 30-day pilot
Every figure above comes from a vendor that sells intent data or services near it. Read them as directional, and see the bias map below.
What is B2B intent data and how is it made?
B2B intent data is evidence that a person or company is researching something. First-party intent comes from your own site, emails and webinars. Third-party intent is bought from vendors who track research activity across other sites. Per Bombora, a surge is inferred from how many people research a topic, relative to normal.
The source decides how much you can trust the signal and how much privacy risk it carries. Knowlee calls first-party data the “GDPR-safer complement” to bought signals.
Steps 1, 2 and 5 follow the Knowlee glossary. Steps 3 and 4 follow Bombora’s own description of its method. Both are vendors describing their own product.
Every step is an inference. The match in step 2 can be wrong, the topic in step 3 can be loose and the surge in step 4 is relative to that company’s normal behaviour. Errors stack.
We use Bombora as the example because it publishes the most about its method. It says its data comes from “a cooperative of B2B publishers”, collected “with consent”. It gives no numeric surge score and does not explain the consent mechanism, so those two questions go on your vendor list.
Why is B2B intent data so tempting to buy?
Because buyers shortlist early. 6sense’s 2025 study of 4,000+ buyers found 94% of buying groups ranked their preferred vendors before first contact. If the shortlist forms before sellers get a call, knowing which accounts are researching has value. That is the best case for B2B intent data.
| Finding | Figure | Source and note |
|---|---|---|
| Buying groups that ranked preferred vendors before first contact | 94% | 6sense Buyer Experience Report 2025 (4,000+ buyers) |
| Deals bought from the top-ranked vendor | 77% | Same report. 6sense’s text elsewhere says 80% |
| Average buying cycle | 10.1 months | Down from 11.3 months, per 6sense via CustomerThink |
| Independent research versus seller engagement | 70/30 to 60/40 | 6sense via Business Wire |
| How far into the process first seller contact happens | 61% | Down from 69%, per 6sense via CustomerThink |
Read these with care. Independent commentary on the study notes that part of the shorter global cycle comes from more physical-goods purchases in the sample. North America was nearly flat, at 11.4 to 11.1 months. The reasons buyers gave for earlier contact are self-reported.
Forrester’s 2025 predictions point the same way. Over half of large B2B deals of US$1 million or more will go through self-serve channels, and half of younger buyers will include ten or more external influencers. Those are predictions, not measurements. Our reading: the shortlist argument is the best case for B2B intent data, but it does not prove that any given feed delivers it.
How accurate is B2B intent data?
No independent, controlled accuracy figure exists in the sources we read. In one DemandScience survey of 750 marketing leaders, only 26% of intent signals converted to qualified opportunities, and 87% said their signals were unreliable or inflated. The publisher sells services, so treat the numbers as directional.
“Intent” signals that convert to qualified opportunities, per a DemandScience survey of marketing leaders.
Self-reported by marketing leaders. It is not a measured conversion rate from our data.The same survey reports that 66% of leaders say campaign metrics often look successful but fail to drive revenue. It also puts the average share of marketing budget wasted at 25%, per respondents. Unclear ROI shows up in about 90% of stacks with 11 to 25 tools, against 62% in stacks of 6 to 10.
The caveats matter. The release does not state the fieldwork dates, and it gives the sample as both “hundreds” and 750. DemandScience sells services that its own findings favour. We use these figures to ask better questions, not to set targets.
Why isn’t a surge the same as a buyer?
A surge shows rising research activity at an account. It does not show who is researching, why, or whether a budget exists. Certification study, competitor benchmarking and general interest can look identical to a purchase in the data. We found six reasons to treat a surge as a hypothesis.
Research is not buying.
Certification study, blog writing and onboarding can look the same as a purchase in the data, per LeadGenius.
Vendor sourceFalse positives waste sales time.
Surges can come from job requirements, competitor benchmarking or general interest, per LeadGenius.
Vendor sourceAmbiguous keywords blur the signal.
Signals aggregated across domains or tied to loose keywords become probabilistic, per LeadGenius.
Vendor sourceAccounts are not people.
Intent flags an account, but sales talks to individuals. You still have to find the right person, and consent rules apply to that person.
ARS reasoning“Above normal” depends on normal.
Bombora measures against an organisation’s usual activity. We have seen no evidence on how that behaves for very large or very small accounts.
ARS reasoningThe outcome evidence is thin.
LeadGenius says six senior leaders told it they saw no gain and did not renew. That is six unnamed anecdotes, so treat it as a claim.
Unverified claimOne fair warning. LeadGenius sells contact data and its own signal products, and its article also says intent stays useful as one signal among many. We include its points because they match how the data is made, not because the source is neutral.
What survives: intent can still rank accounts for human attention. It is a poor trigger for automatic outreach.
Who is saying what about B2B intent data?
Most B2B intent data statistics come from companies that sell intent data, intent platforms or marketing services. That does not make them false. It means the publisher benefits when you believe them. Ask what a source sells before you quote it, and badge every figure by who produced it.
| Source | Sells | Says | Read it as | Type |
|---|---|---|---|---|
| 6sense Buyer Experience Report 2025 | Intent and ABM platform | 94% rank vendors before contact | Large sample, but vendor-run | Vendor |
| DemandScience State of Performance Marketing 2026 | B2B performance marketing | 87% unreliable signals; 26% convert | 750 leaders, run by TrendCandy. Fieldwork dates not stated | Vendor |
| Bombora, “Our data” | Intent data | 5,000+ sites, 15.8B interactions a month | Marketing claims, unverified | Vendor |
| LeadGenius, “Is intent data BS?” | Contact data and competing signals | Intent often wrong; six leaders did not renew | Six unnamed anecdotes from a competitor | Unverified |
| Knowlee glossary | AI sales and signal tools | GDPR Article 6 risks of third-party data | Its own legal view, not legal advice | Vendor |
| Forrester Predictions 2025 | Research and advisory | Self-serve share, 10+ influencers | Predictions, not measurements | Independent |
| UK ICO guidance | Regulator | Legitimate interests three-part test | Primary guidance | Regulator |
Our own bias belongs on the table too. ARS B2B Social Bridge is a lead generation business. We sell neither intent data nor compliance advice, but we benefit when marketers care about lead quality. We have tried to keep our view separate from the sourced numbers.
Can you act on B2B intent data without breaking privacy rules?
Only if you treat the signal as a reason to prioritise, not as consent to message. The UK ICO says direct marketing can be a legitimate interest, but it must pass purpose, necessity and balancing steps. In Germany, per Overloop, intent signals do not replace prior express consent for email.
Third-party intent starts with data about people, even when it is sold as company data. Knowlee argues that consent given for advertising personalisation may not cover selling research activity for outbound targeting, and that a buyer of the data needs its own documented basis. That is its own characterisation, not legal advice.
A safer default, in our reading: use third-party intent to prioritise accounts and shape content. Use your own first-party and consented data to reach individuals. Knowlee suggests the same approach, and it fits the ICO’s necessity and balancing steps. Country rules for email, calls and lead sharing sit in our B2B email compliance atlas.
The signal-to-action ladder. Not every signal deserves a message, so climb only as far as the evidence and the rules allow. This ladder is ARS reasoning, not a published standard. ARS reasoning
- Watch. Log the surge. Do nothing yet. See how often it repeats for the same account.
- Prioritise. Move the account up the list for human research and account planning.
- Tailor. Match ads, content and landing pages to the surging topic.Check ad-targeting rules
- Warm. Send relevant content to known contacts who have already opted in.Consent on file
- Reach out. Contact named individuals directly, only where the market’s rule is met.Atlas card first
A surge is a guess. Low rungs cost little if the guess is wrong, while high rungs cost trust and can break consent rules.
How do you test B2B intent data before you renew?
Ask the vendor ten questions in writing, then run a 30-day matched-group pilot. Compare accounts that show a surge with similar accounts that do not, handled by the same reps with the same offer. Decide on one metric and write your pass mark before the pilot starts.
Ten questions for your vendor
- What exactly counts as a surge, and how is the score calculated?
- How many people at an account must show activity before it flags?
- Over what time window, and compared with what baseline?
- How do you match visitors to companies, and what is your match error rate?
- What consent was given, to whom, and can I see the wording?
- Is any signal repackaged from another vendor or from first-party data?
- How often is each topic refreshed, and how old can a signal be?
- Can you show a result that was measured against a holdout?
- What happens to my data and spend if I leave after year one?
- Can I run a paid 30-day pilot with a clean exit?
If a vendor cannot answer these in writing, you are buying on trust. A case study with a pipeline number and no holdout is a story, not a test.
A 30-day pilot you can run
- Define. Pick one topic and one segment. Choose one success metric: meeting rate, SQL rate or cost per SQL. Freeze your baseline and write your pass mark now.
- Split. Match accounts by size and industry. Half get signal-led attention. Half get your normal process.
- Run and log. Reps note whether the contact recognised the topic. Check consent rules before each message.
- Read out. Compare the two groups on your metric. Renew, shrink or stop, then share the result with your vendor.
We do not quote an industry pass mark, because we found no independently sourced one. Set yours from your own baseline. A matched control group tells you whether the signal helped or whether those accounts would have replied anyway.
This pilot is ARS reasoning. To score what comes out of it, pair B2B intent data with the lead scoring model in our B2B lead quality report.
What do critics say about this view of B2B intent data?
Vendors say intent data produces large pipeline gains, and many publish case studies to show it. Critics of that evidence note that the gains are often attributed by the client, with no holdout. We found no independent controlled trial, so we cannot say the gains are false or real.
Vendors will say a 26% conversion rate is a healthy return for a top-of-funnel signal, and that the shortlist data supports buying early. That is a fair argument. The 6sense numbers do show buyers deciding before they talk to sellers.
The Forrester items are predictions, not measurements. The DemandScience and 6sense figures are self-reported by respondents, and the vendors behind them sell the products in question.
This paper has limits too. It leaves out pricing, which we could not source independently. It does not cover review-site or contact-level intent products, and it could not verify Bombora’s scale and consent claims, DemandScience’s fieldwork dates, 6sense’s method or LeadGenius’s six interviews. Gartner buyer-journey figures were left out because we could not verify them.
What do marketers ask most about B2B intent data?
What is B2B intent data?
B2B intent data is evidence that a company or person is researching a topic. First-party intent comes from your own website, emails and webinars. Third-party intent comes from vendors who track research activity across other sites. Either way, it is an inference about interest, not a confirmed plan to buy.
How accurate is B2B intent data?
Nobody has published an independent, controlled accuracy figure. In a DemandScience survey of 750 marketing leaders, only 26% of intent signals converted to qualified opportunities, and 87% called their intent signals unreliable or inflated. DemandScience sells services, so treat those numbers as directional and test on your own pipeline.
What is the difference between first-party and third-party intent data?
First-party intent comes from properties you control, such as your site, emails and product. Third-party intent is bought from vendors who aggregate research activity across publisher sites, review platforms and content networks. Per Knowlee, first-party data is the GDPR-safer complement. Third-party signals usually reach you as a company, topic and score.
Is B2B intent data legal under GDPR?
It depends on how the data is collected and used. The UK ICO says direct marketing can be a legitimate interest, but it must pass a three-part test of purpose, necessity and balancing. A signal does not replace consent where the channel rules require it. This is general information, not legal advice.
How do you test whether B2B intent data works?
Run a 30-day matched-group pilot. Pick one topic and one segment, then split similar accounts in half. One half gets signal-led attention and the other gets your normal process, with the same reps and offer. Compare meeting rate, SQL rate or cost per SQL against a pass mark you set beforehand.
Is the B2B Intent Data Reality Check free to download?
Yes, the B2B Intent Data Reality Check 2026 from ARS B2B Social Bridge is a free 16-page PDF. It has a source bias map, ten vendor questions, a signal-to-action ladder and a 30-day pilot plan. Use the download button on this page. It is general information, not legal or investment advice.
What should you do next about B2B intent data?
Do not cancel intent data and do not trust it blindly. Download the report, send the ten questions to your vendor and run the 30-day pilot with a holdout. Keep consent rules in front of every message that follows a signal, and renew only on your own evidence.
Treat a surge as a hypothesis about an account. Judge B2B intent data the way you would any media spend: ask for a test, read the contract exit and keep proof of what it changed.
For wider business and marketing analysis, read The Business Perspective. For technology and automation coverage that touches outreach tooling, see IoT Mail Bridge and IoT Insights Hub. ARS B2B Social Bridge publishes more research at arsb2bsocialbridge.com.
Get the free B2B Intent Data Reality Check 2026 from ARS B2B Social Bridge. Sixteen pages, a source bias map, ten vendor questions and a 30-day pilot.
Download the free PDF