Intent Data Playbook: Why It’s Falling Apart

Walk into nearly any B2B marketing conference today, and you’ll hear the same pitch. Someone will mention intent data, talk about surge topics and in-market accounts, and show off a slick dashboard. They’ll name their provider , maybe a stack of layered tools. It all looks impressive.

But here’s the uncomfortable truth: a massive portion of that data originates from the same well. Your three closest competitors are drinking from it, too. This is intent data commoditization, and it’s the reason your cost per lead (CPL) keeps climbing while your conversion rates flatline.

Most third-party intent data falls into a few familiar buckets: publisher co-ops (networks of B2B sites sharing anonymized content consumption data), software review platforms like G2 and TrustRadius, and bid stream data. That last one is the biggest and least understood.

Bid stream data comes from programmatic ad exchanges. Every time a webpage loads an ad, metadata about the visitor , including the content they’re viewing , flows through ad exchange bid requests. Some intent providers tap this stream to infer topic interest. The scale is enormous, with billions of signals daily.

The bid stream problem

At that scale, tradeoffs become glaring. Accuracy is lower than co-op or editorial data. Resolution is mostly account-level (IP-resolved to companies). And the privacy footing is shaky. As one provider candidly admits, “B2B intent data providers relying on bidstream stand on very weak ground” under GDPR, because real-time bidding mass collection often happens without meaningful consent.

Some providers are even more transparent about their data lineage. They tell prospective customers their intent data delivers breadth thanks to “direct access to the bidstream , the source for the most intent signals.”

At least they’re honest. But the structural issue is clear: when one source supplies the largest share of an industry’s intent, and most platforms either tap that source directly or resell from a handful of co-ops, you end up with a market where everyone sees the same signals at roughly the same time.

The downstream economics are predictable. According to DemandScience’s 2026 State of Performance Marketing Report, 87% of organizations say their marketing investments produce unreliable or inflated intent signals, and only 26% of those signals convert into qualified opportunities. Two-thirds of leaders say their campaign metrics often look successful but fail to drive revenue.

If you and three competitors are bidding on the same surging accounts simultaneously, you’re just bidding up the price of a meeting.

What signal convergence should actually mean

Let’s untangle two ideas that often get confused.

Intent data commoditization is the problem above: same data, same accounts, same plays, higher costs.

Signal convergence is the opposite. It’s the moment when your account-level signals , firmographic fit, technographic match, hiring patterns, funding , and your contact-level signals , a real human at that account engaging with your content, your competitors, and your category , intersect. That intersection is where marketing and sales activities converge and produce meetings.

You can’t reach that intersection by buying the same packaged feeds as everyone else. You need to build a richer signal layer, much of it from sources your competitors aren’t using.

That’s custom signal capture. And it’s accessible even for short-staffed marketing teams.

Custom signal scraping: A primer

Instead of , or in addition to , buying packaged intent feeds, you systematically capture buying signals from public sources that aren’t already in someone else’s data co-op. Done well, you end up with a proprietary signal your competitors literally can’t purchase.

Here are the categories worth considering:

• Job postings: A company posting a senior security role is a signal. Posting five in a quarter alongside a new VP of Engineering? That’s a much louder signal. Job descriptions also leak technology stack, team structure, and strategic priorities.The combination is what matters. One signal is noise. A funding round + three engineering hires + a new CIO + a podcast appearance where the new CIO mentions the exact problem you solve , that’s a signal to go get a meeting.

Data comes in all shapes and sizes

The right approach depends on your time, budget, and operational maturity. Here’s how I’d think about it for three different team sizes.

Small teams (Seed to Series A, marketing team of 1-5)

You can’t afford a $100,000 Demandbase or 6sense contract, and honestly, you shouldn’t. At this stage, the goal is precision over scale.

Start with a focused list of 100-300 target accounts. Build your own signal layer using:

• A workflow tool like Clay to orchestrate enrichment, scraping, and routing into your CRM.Total spend can land well under $2,000 a month. The output is a richer, more current view of your top 200 accounts than most enterprise teams have on theirs. You catch buyers your competitors can’t see, and you reach them while the signal is still warm.Mid-market teams (Series B to D, marketing team of 6-10)You have more budget, more accounts to cover, and more pressure to show predictable pipeline. A two-tiered approach works here.For your top tier (the 200-500 accounts you actually want to win), run the small-team playbook above with more rigor. Document signal definitions. Build automated scoring. Connect everything to a dashboard your sales team checks daily.For the broader audience (5,000+ accounts), you can layer in one packaged third-party source, but pick one that’s differentiated from what your competitors are likely buying.G2 Buyer Intent and TrustRadius give you software comparison signals you can’t reproduce. UserGems gives you job-change signals at scale. A category-specific provider , like Onfire for technical buyers , often beats a generic feed for less money.Enterprise teams (Marketing team of 10+)This is where I see the most waste. Big intent contracts get auto-renewed. The data flows into a system nobody really audits. Sales reps complain that “the intent data isn’t useful,” which is sometimes a skill problem, and sometimes a real problem.Two moves matter at this scale.
• Audit what you’re actually getting from your packaged providers: Run an overlap analysis. If your three providers are giving you 70% the same accounts, consolidate or replace one with something genuinely additive. Custom-scraped signals , hiring, technographic, community, leadership change, and dark social , almost always provide higher uniqueness than another bid-stream-derived feed.

Where to start after you finish reading this

If you take one thing from this, let it be this: stop buying signals and start building them.

A practical first step: pick 50 target accounts. Spend two hours setting up a simple weekly scrape across three sources you don’t currently monitor , say, job postings, leadership announcements, and podcast appearances. See how many of those 50 accounts produce a real signal in 30 days.

Buying intent data isn’t going away. You’ll probably still need a packaged provider or two. But the competitive edge in 2026 won’t come from spending more on the same data your competitors already have. It’s going to come from the signals you build yourself, and the ones nobody else thought to look for.