Intent data providers: the buyer's guide

Your BDRs have the flagged accounts. Bombora says 40 of them are surging on category terms. The intent platform marks them high-priority.

They open the CRM. Decision-maker mobile coverage: 12%. The remaining contacts are generic office lines, LinkedIn profiles with no activity, and a few records that haven't been refreshed since 2023.

Intent without a reach layer isn't pipeline. It's a prioritized list you can't action.

That's the structural problem. And it's not a vendor quality issue. Bombora, 6sense, and G2 generate real signal. The gap is architectural: intent platforms tell you who's in-market, not how to reach them. For enterprise and mid-market SaaS ICPs with strong LinkedIn presence, a ZoomInfo or Cognism layer closes that gap. For restaurant operators, franchise networks, HVAC contractors, and any non-LinkedIn-native segment - it doesn't. ~50% of local decision-makers have no LinkedIn profile. The standard contact layer returns 10–20% mobile coverage. The intent signal fires on accounts you can't reach.

This guide covers the four architectural types of intent providers, an honest vendor-by-vendor breakdown, how to run a structured evaluation, and the coverage gap that kills intent programs targeting local and SMB segments.

For the educational primer on signal types and workflows, read B2B intent data before you score vendors in the sections below.

1. What intent data actually is (and what it isn't)

Most buyers evaluating intent data providers are actually conflating three separate categories of data, and the confusion leads to mismatched purchases. Intent data, contact data, and account intelligence are different layers of the targeting stack, and intent sits on top of the others.

1.1. Intent data - signal about account buying behavior

Intent data is behavioral evidence that an account is actively researching a category, solution, or competitor. It comes in three forms, each with different coverage characteristics and fidelity levels.

Third-party intent is the largest category: publisher-consortium behavioral signals collected from networks of B2B sites (Bombora's co-op spans 5,500+ publishers) and licensed to buyers and to other platforms. When a company's employees are reading comparison articles, vendor reviews, and category explainers across the publisher network, those signals aggregate into account-level "surge" scores. Second-party intent comes from review and comparison platforms - G2's Buyer Intent flags users actively researching your category or comparing you against competitors. First-party intent is narrower but highest-fidelity: reverse-IP identification of anonymous visitors to your own site, surfacing which accounts are interacting with your owned content without converting.

Each type answers a different question about an account's buying stage. Third-party intent answers "who is researching this category across the internet." Second-party answers "who is comparing vendors right now." First-party answers "who is on your site."

What intent data is not

Intent data is not contact data. Intent providers identify which accounts are in-market. They don't return the decision-maker's mobile numbers, direct email, or job title at that account. The two layers serve different functions and require different vendors.

Intent data is also not firmographics. Knowing an account's revenue, employee count, and tech stack tells you whether it fits your ICP. Knowing it has high intent signals tells you it's actively buying. Neither replaces the other. The complete targeting stack is: ICP filtering (firmographics + technographics) + prioritization (intent signals) + outreach (contact data). Teams that buy intent data expecting it to replace list-building miss the architecture entirely.

1.2. The segment qualifier

Intent data value scales with account volume, buying committee size, and the density of that ICP within the publisher co-op's network. Enterprise ABM programs running against 500+ named accounts in B2B tech get the most value. Their target accounts are well-indexed by publisher networks, and the signals are dense enough to be meaningful. SMB teams running high-volume outbound against 10,000+ targets get diminishing returns as intent signals thin out. For local business ABM (restaurants, HVAC contractors, franchise operators, multi-unit retail), third-party intent coverage is structurally sparse. These decision-makers don't read G2 or TechCrunch. The publisher co-ops that generate intent signals don't index their behavior the way they index corporate B2B buyers.

2. The four types of intent data providers

The intent data market is architecturally stratified. Understanding the underlying model before evaluating vendors saves significant time and budget.

2.1. Third-party intent co-ops (Bombora-style)

Bombora pioneered the publisher-consortium model. A network of 5,500+ B2B publishers contribute anonymous behavioral signals, which Bombora aggregates into account-level surge scores and licenses to buyers. This model is also embedded by most other vendors: ZoomInfo, Cognism, Lead411, and many CRM and ABM platforms license Bombora as their underlying intent layer rather than building proprietary signal networks. For a team that wants raw intent data without an ABM platform wrapper, Bombora direct is the canonical source. Strongest on tech buyers and enterprise B2B; structurally weaker on non-tech verticals and local business segments where the publisher network has thin indexing.

2.2. Predictive intent platforms (6sense, Demandbase)

6sense and Demandbase are ABM platforms with integrated predictive intent - they model buying stage probability on top of raw intent signals rather than passing through raw data. The output isn't a surge score but a buying stage classification: "awareness," "consideration," "decision." This is more useful for teams that can act on stage-specific recommendations and have the analyst capacity to operationalize the model's outputs. Predictive models depend on training data; effective for accounts that resemble the training set, less effective for outlier segments or novel ICPs.

2.3. First-party intent / reverse-IP identification

Platforms like Dreamdata, RB2B, and Lead Forensics identify anonymous website visitors by company via IP lookup and surface account-level visitor behavior. The use case is narrower than third-party intent. It only catches accounts that visit your site. But the fidelity is higher because the signal is direct behavioral evidence rather than inferred from publisher network activity. Most useful for teams with meaningful inbound traffic and an attribution need alongside the intent function.

2.4. Review + comparison site intent (G2)

G2's Buyer Intent flags users researching your category or comparing you against competitors on the G2 platform. This is second-party intent. Not behavioral signals spanning the whole web, but high-fidelity evidence of active vendor evaluation. Users on G2 are further along the buying journey than users triggering third-party intent signals. For teams with strong category presence on review sites, G2 Buyer Intent is often the first intent product added because the signal is immediately legible: "someone compared us to Competitor X this week."

3. The architectural question intent buyers skip

Intent data surfaces which accounts are in-market. That signal is useful only if the team can actually reach the decision-makers at those accounts. And whether that's possible depends on the contact data layer, not the intent layer.

3.1. Intent without reach is waste

A BDR team working 100 high-intent accounts needs contactable decision-makers at those accounts. If the contact database returns 10–20% decision-maker mobile coverage - which is the standard ceiling for traditional providers (ZoomInfo, Apollo, Clay, Cognism, Lusha) when the ICP includes local or non-LinkedIn-native contacts - 80% of intent-flagged accounts become ghost entries. Known to be in-market. Unreachable via outbound.

This is the DQ cascade that breaks intent programs in practice: the intent layer surfaces 100 accounts, the contact enrichment step returns valid decision-maker mobiles on 12–18 of them, and the BDR works those 12–18 while the other 82 sit in a "researching" queue that never gets actioned. The 40% of BDR capacity that goes to manual research. At $100–120K per rep per year, that's $40–50K per rep per year in manual enrichment overhead, represents the gap between the intent layer's promise and what the contact layer can actually deliver (per industry compensation benchmarks).

3.2. What this means for local/smb ABM

For teams selling into local businesses, trades operators, or franchise decision-makers, the breaking point in the intent stack is usually the contact layer, not the intent layer. Third-party publisher co-ops don't index local business buying behavior at the same depth as corporate B2B. Running Cognism or ZoomInfo enrichment against a restaurant or franchise account list will return thin intent coverage and thin contact coverage simultaneously - because the same LinkedIn-dependency architecture underlies both.

The structural fix for local ABM is a two-layer approach: pair a traditional intent provider (Bombora, 6sense) for the LinkedIn-native accounts in your mix with a discovery-first contact data layer for the local and SMB tail those providers miss. DataLane, a U.S.-only data layer purpose-built for local and SMB segments, provides 60%+ decision-maker mobile coverage at 80%+ accuracy on local business accounts. The 17M+ location dataset sources from state licensing boards, permit filings, and franchise registries rather than LinkedIn, which is why it covers segments that traditional enrichment providers structurally cannot. DataLane is not an intent data provider. It's the contact infrastructure that makes intent-flagged local accounts actually reachable.

3.3. When intent data doesn't move the needle

There are four scenarios where intent data reliably underdelivers. Brand-new categories with no review-site presence: G2 intent has nothing to report, and third-party signals are thin because buyers are searching, not comparing. Local business verticals: publisher co-op coverage is structurally sparse. Early-stage startups with sub-100-account TAMs: intent signals add noise to a universe too small to benefit from prioritization. Teams with thin contact data on flagged accounts: intent becomes a pipeline illusion. A growing queue of accounts the team knows are in-market but can't reach. The fourth scenario is the one most intent programs run into without naming it.

4. Intent data provider landscape

The intent data market has three category leaders (Bombora, 6sense, G2), two integrated ABM platforms with embedded intent (Demandbase, ZoomInfo's intent add-on), a first-party specialist (Dreamdata), and a second tier of challengers. Here's an honest breakdown of what each actually delivers.

4.1. Bombora - the third-party intent standard

Bombora is the infrastructure layer most of the intent data category is built on. The Company Surge® product aggregates signals across 5,500+ B2B publishers into account-level intent scores by topic. If you want to know which accounts are researching "contract lifecycle management" or "field service software" across the B2B web, Bombora is the canonical source.

The critical thing to understand about Bombora is how widely its data is licensed downstream. ZoomInfo, Cognism, Lead411, and dozens of CRM and ABM platforms embed Bombora as their intent layer rather than building proprietary signal networks. This means many buyers are already running on Bombora data without knowing it, through their existing contact platform's "intent" add-on. Buying Bombora direct gives the same underlying signal with more control over the activation layer.

Bombora's structural limitation is the publisher co-op's composition. The network indexes corporate B2B tech buyers well. Coverage thins significantly for non-tech verticals, local business operators, and franchise decision-makers who don't read the B2B publishers in the co-op. If your ICP is mid-market SaaS or enterprise software buyers, Bombora is the most defensible intent source. If your ICP includes significant local or vertical-specific accounts, run a coverage test on your actual accounts before committing.

Pricing is custom and typically starts at $30K annually for direct access, scaling with topic volume, account volume, and seat count. Integration is available into most major CRM, marketing automation, and ABM platforms.

4.2. 6sense - predictive intent + ABM orchestration

6sense is not a raw intent data provider. It's an ABM platform with integrated predictive intent, and the distinction matters for how you evaluate it. Rather than returning account surge scores, 6sense runs a predictive model across first-party, third-party (Bombora), technographic, and firmographic signals to output buying stage classifications: which accounts are in "awareness," "consideration," or "decision" stage.

The value proposition over raw Bombora data is the model layer: instead of interpreting surge scores yourself, 6sense tells you what stage each account is in and recommends stage-appropriate outreach. For enterprise GTM teams with the analyst maturity to operationalize stage-specific playbooks, this is a meaningful capability uplift. For teams that don't have defined stage-specific sequences, the model layer adds cost without adding workflow change.

6sense pricing is enterprise-only and typically six figures annually. Deployment requires dedicated RevOps or marketing ops resources to configure account scoring, integrate with CRM and SEP, and maintain the model's account universe. It's not a tool you plug in and get immediate value from. It rewards teams that can act on its recommendations at scale. The intent layer underlying 6sense is shared with Bombora; the differentiation is the model, orchestration, and advertising activation on top of it.

One note on local/SMB ABM: 6sense's predictive models are trained on data that heavily skews toward enterprise and corporate B2B accounts. Sending 1,000 local business accounts through 6sense and expecting intent enrichment back on 1,000 of them is not realistic, enrichment rates on local segments run significantly lower (in some cases, 10% or less) because the training data and signal network don't index those accounts at the same depth. Run a coverage test with your actual ICP before committing at 6sense's price point.

4.3. G2 buyer intent, review + comparison intent

G2 Buyer Intent is the clearest intent signal in the category because users generating it are actively evaluating vendors. Someone on G2 comparing your product to two competitors is much further along the buying journey than someone triggering a Bombora surge score by reading category content. For teams with meaningful G2 presence, established review volume, category visibility, G2 Buyer Intent is often the first intent product purchased because the signal is immediately interpretable and actionable.

The limitation is scope. G2 intent only covers buyers actively on G2. It misses accounts researching on other review platforms (TrustRadius, Capterra), accounts doing early-stage research via publisher content (covered by Bombora), and any ICP that doesn't use G2 as part of its vendor evaluation process. Local business decision-makers and franchise operators are largely absent from G2's buyer population.

G2 Buyer Intent is available through G2's Track and Pro subscription tiers, typically priced at $15K–$50K annually depending on category depth and alert configuration. It integrates with Salesforce, HubSpot, Marketo, and Outreach. For most teams, G2 intent is a complement to Bombora or 6sense rather than a standalone, because each covers a different behavioral window in the buying journey.

4.4. Cognism (intent + contact bundle)

Cognism embeds Bombora intent signals within its contact data platform, so teams can buy contact enrichment and intent data in a single contract. The bundling is pragmatic for teams that want to avoid managing separate vendor relationships, and Cognism's contact coverage is strong in EMEA and solid in North America for LinkedIn-native ICPs (400M+ contacts, Elevate® verified mobiles, strong corporate and mid-market depth).

The architectural caveat applies here the same as with all five of the LinkedIn-dependent providers, ZoomInfo, Apollo, Clay, Cognism, and Lusha all source primarily from LinkedIn scraping plus corporate web data. For enterprise and corporate B2B, this architecture works cleanly. For local business, SMB, and non-LinkedIn-native segments, the same structural ceiling applies regardless of which bundled provider is used. The intent signals within Cognism's platform are Bombora signals; switching from Bombora direct to Cognism for the intent layer produces the same underlying data.

4.5. Demandbase - ABM platform with integrated intent

Demandbase is the other category leader alongside 6sense in the ABM platform space, bundling intent data, account intelligence, account-based advertising, and orchestration in a unified platform. For this intent-specific evaluation: Demandbase intent is integrated and not purchasable separately. Teams evaluating Demandbase should review it as an ABM platform investment, not as a standalone intent data purchase. Pricing is enterprise-tier, comparable to 6sense.

4.6. Dreamdata - first-party intent + attribution

Dreamdata combines first-party intent identification (reverse-IP company lookup for anonymous site visitors) with B2B revenue attribution. The product answers two questions simultaneously: which companies are interacting with our site, and what touchpoints drove closed deals. This is a narrower scope than third-party intent. It only covers accounts that visit owned channels. But the fidelity is high because the signal is direct behavioral evidence, not inferred from publisher activity.

For teams with meaningful inbound traffic and an attribution problem to solve, Dreamdata is worth evaluating alongside a third-party intent provider rather than instead of one. Pricing ranges from roughly $30K to $60K annually for mid-market tiers, scaling with account volume and attribution complexity.

4.7. ZoomInfo (intent add-on)

ZoomInfo offers a Bombora-powered intent add-on to its core contact platform. For teams already on ZoomInfo enterprise contracts, this is the path of least resistance to adding intent signals. Same underlying Bombora data, bundled into existing contract and interface. There's no architectural differentiation between ZoomInfo intent and Bombora direct beyond the bundling. Teams that are already ZoomInfo-committed and want intent without a second vendor relationship will default here. Teams that are not already committed to ZoomInfo should evaluate Bombora direct against the full ZoomInfo bundle on unit economics before deciding.

4.8. Foundry intent, lead forensics, intentsify, lead onion

These four vendors appear regularly in intent data listicles and compete primarily in the mid-market segment below the Bombora/6sense tier. Foundry (formerly IDG) offers publisher-network intent similar to Bombora, with strength in tech buyer segments. Lead Forensics is a legacy reverse-IP identification player with long market tenure and a user base that has largely stayed put rather than actively re-evaluated. Intentsify offers proprietary intent signal activation and ABM workflow tools. Lead Onion packages intent signals with contact data in a mid-market bundle.

None of the four has the category dominance or ecosystem integration depth of Bombora, 6sense, or G2. Teams evaluating them should run the same coverage test methodology against their actual ICP and compare return rates directly against the category leaders before committing. The pricing advantage in the mid-market tier is real; the question is whether the coverage rates justify replacing a Bombora integration that may already exist within a current platform contract.

4.9. Why you don't see DataLane in this list

DataLane is not an intent data provider. And the distinction matters architecturally. DataLane is the contact data layer that sits underneath an intent data stack for local, SMB, and franchise segments. Intent data identifies which accounts are in-market. DataLane provides the verified decision-maker mobile contacts that make those flagged accounts actually reachable for outbound.

For teams running ABM motions that include any volume of local business, independent contractor, franchise operator, or non-LinkedIn-native accounts, DataLane is the infrastructure addition that closes the contact coverage gap the intent layer exposes. The 60%+ decision-maker mobile coverage on local/SMB accounts (17M+ U.S. locations, sourced from licensing boards, permit registries, and franchise filings rather than LinkedIn) addresses the segment that ZoomInfo, Apollo, Clay, Cognism, and Lusha structurally cannot reach. DataLane is U.S.-only, batch-delivery. It is not an intent data platform. Pair it with Bombora or 6sense for the intent layer; DataLane handles the contact layer underneath.

5. How to evaluate an intent data provider

Most intent data evaluations are rigged in favor of the vendor before the first demo. The methodology below corrects for that.

5.1. Start with your ICP and motion

Enterprise ABM, mid-market outbound, and local business ABM have fundamentally different intent data needs. Enterprise ABM against named accounts benefits most from predictive models (6sense, Demandbase) because the account universe is small enough to act on stage-specific recommendations. Mid-market outbound at volume benefits from third-party intent signals (Bombora) as a prioritization filter on a larger target universe. Local business ABM benefits from vertical-specific intent signals (licensing events, permit filings, tech stack changes) rather than publisher co-op signals. Match provider architecture to motion before evaluating vendor feature matrices.

5.2. Audit the signal source

Every intent vendor should be able to tell you: what is the underlying data source? Publisher co-op (Bombora, Foundry)? Proprietary signal network (6sense)? Review platform (G2)? First-party web (Dreamdata, RB2B)? The answer determines coverage characteristics, dwell biases, and what segment the signals are actually indexing. A vendor that can't clearly explain its signal source is a vendor whose data you can't evaluate.

5.3. Coverage on your target accounts

Send the vendor 100 accounts from your real ICP. Not a vendor-curated sample, your actual accounts. And ask what percentage have intent signals in the last 90 days. Compare that return rate to your other vendors' coverage on the same 100 accounts. Intent providers without segment-specific coverage return noise, not pipeline. A vendor that performs well on vendor-selected accounts and poorly on your actual accounts has a mismatch between their pitch and your segment. The bake-off test must use your list, not theirs.

Provider database size doesn't tell you intent signal coverage for your segment. A platform indexing 50M corporate web visits and 30M B2B contacts has near-zero coverage for local business operators, who aren't doing corporate research or LinkedIn activity. Test your specific ICP. The return rate on your 100 accounts is the only number that tells you whether the platform covers what you're actually targeting.

5.4. Integration with your existing stack

Intent data sitting in a standalone dashboard that doesn't push into Salesforce or HubSpot is wasted within 90 days. Before evaluating any intent provider, verify: native CRM connector, ABM platform integration (if applicable), marketing automation hooks (Marketo, HubSpot), sales engagement platform integration (Outreach, Salesloft), and advertising audience integration (LinkedIn, Google, Meta). Stack integration determines whether intent signals reach the BDRs and marketers who act on them; without it, the data never leaves the platform.

5.5. Intent signal recency and refresh cadence

Weekly refresh is the industry standard for mid-market intent products. Enterprise-tier providers typically offer daily refresh. Stale intent signals are worse than no intent signals in one specific way: a team working accounts that were flagged as "in buying stage" four weeks ago may be working accounts that already selected a vendor. Verify refresh cadence before signing, and verify it contractually, "weekly" in the pitch deck sometimes means "bi-weekly" in the contract.

5.6. Don't buy intent without a plan to action it

The single most common intent data failure mode is buying the signal without defining the response workflow. Before signing any intent contract, your team needs answers to: which BDRs are assigned to intent-flagged accounts, what is the SLA from flag to first touch, what sequence do they run, what does "intent-worked" mean in CRM fields, and how will intent-influenced pipeline be tracked separately from general outbound? Without a defined workflow, intent data becomes a growing queue that no one systematically works, which produces no ROI and an easy cancellation conversation at renewal.

5.7. The two bake-off traps

Two evaluation mistakes consistently skew intent vendor comparisons in the vendor's favor. The first is accepting vendor-selected account samples. Never let the vendor choose the accounts for the coverage test. Send your list, measure their return. The second is correlation-as-causation attribution: vendors frequently present data showing "accounts with intent signals closed at 2–3x the rate of accounts without signals." Read the methodology carefully. In-market accounts are over-represented in both the intent-signal population and the closed-won population by definition. The relevant question is whether intent-signal prioritization caused faster pipeline velocity or higher close rates. Not whether accounts that were already buying (and therefore already generating intent signals) closed at higher rates.

6. Intent data for non-LinkedIn-native ICPs

The segment most buyers' guides skip is the one where intent data most reliably underperforms its pitch.

6.1. The coverage gap in local and vertical ICPs

Third-party publisher co-ops (Bombora, Foundry) index buying behavior on B2B tech publisher sites. Their signal networks are strong for enterprise SaaS buyers, corporate procurement teams, and mid-market technology purchasers, segments that read G2, attend virtual summits, and interact with B2B tech publisher content. For local business operators, restaurant owners, HVAC contractors, multi-unit franchise decision-makers, independent trades businesses. The publisher network has thin indexing. These operators aren't reading TechCrunch or attending SaaS webinars. Their buying behavior doesn't generate the publisher signals that Bombora aggregates. The data gap isn't a data quality issue; it's an architectural mismatch between the signal source and the segment.

6.2. Vertical-specific intent sources

For local business ABM, the meaningful intent signals are structurally different from corporate B2B intent signals. New business licensing events (a new contractor license filed in a state registry) indicate a business entering growth mode. Permit filings (HVAC, electrical, plumbing permits at above-baseline rates) signal capacity expansion. Franchise disclosure updates indicate new-unit openings. POS system changes serve as tech stack displacement signals. A business that recently switched point-of-sale systems is mid-change-cycle and potentially open to adjacent software purchases. Some of these signals are surfaced by discovery-first data providers that source from licensing boards and permit registries rather than publisher networks. Traditional third-party intent providers don't capture these events because they're not behavioral web signals, they're public record events.

6.3. The two-layer stack for local ABM

For GTM teams running mixed ICPs. Some enterprise or corporate accounts alongside local or SMB accounts; the right architecture is two layers, not one universal provider. A traditional intent provider (Bombora, 6sense) handles the LinkedIn-native portion of the account universe where publisher signal coverage is dense. A discovery-first contact data layer (DataLane) handles the local and SMB tail that intent providers structurally miss, providing the decision-maker mobile contacts that make intent-flagged local accounts actionable even when the intent signal itself is thin. The two layers serve different segments; neither replaces the other.

7. Budget and ROI

Intent data budget decisions made in isolation routinely produce sticker shock when the full stack cost becomes apparent.

7.1. What you'll actually pay

Entry-tier intent products (Dreamdata, smaller first-party providers, Lead Forensics) start around $15K–$30K annually. Mid-market tiers - Cognism's intent bundle, G2 Buyer Intent at category depth, Foundry intent. Typically run $30K–$80K annually. Enterprise intent data (6sense, Demandbase, Bombora direct at scale) ranges from $80K to $250K+ annually depending on seat count, account volume, topic depth, and platform tier. Custom pricing is universal above the entry tier; published pricing should be treated as a floor, not a ceiling.

7.2. Measuring ROI honestly

The right ROI question for intent data is not "did intent-flagged accounts close at higher rates?" That metric is confounded by the fact that in-market accounts are over-represented in both the intent-signal population and the closed-won population. The honest ROI question is: did intent-signal prioritization produce faster pipeline velocity or account coverage that wouldn't have happened without the signal? That requires measuring account entry rate (what percentage of intent-flagged accounts received a first touch within the SLA window), pipeline contribution from intent-first account entries, and conversion rates compared to non-intent-triggered accounts with equivalent ICP scores. Without that attribution stack, ROI measurement is a post-hoc rationalization, not a causal claim.

7.3. The stack cost reality

Intent data is one line item in a stack where every layer has a budget. A fully-loaded intent-driven GTM stack typically includes: intent data provider ($30K–$250K), contact data provider ($15K–$60K+), ABM platform if separate from intent ($30K–$200K), CRM, marketing automation, sales engagement platform, and ad spend for retargeting. Teams evaluating intent data in isolation from the full stack often sign the intent contract and discover the contact data layer can't cover the flagged accounts, or that the ABM platform isn't included in the intent vendor's contract, or that the workflow to action the signals requires a RevOps investment that wasn't budgeted. Price the workflow, not just the data.

Frequently asked questions

Data quality compounds. Fix the source layer first; the workflow layer is downstream.