When the Data Disagrees With You

Shashi Bellamkonda  |  misunderstoodmarketing.com

I have spent most of my career believing that relationships win in B2B marketing. That persistence signals commitment. That a good salesperson who stays present in an account, even a quiet one, is doing the right thing. Twenty-five years as a practitioner, including time as a chief marketing officer, reinforced that belief every time a deal closed after months of patient follow-up.

So when I read the new research from Labs by Demandbase, my first instinct was to push back. The finding that increasing activity on stalled accounts actually lowers conversion rates cuts against something I have watched work, or believed I watched work, for a long time. But the honest response to data you disagree with is not to dismiss it. It is to sit with it.

Here is what changed my mind, at least partially, and what I think practitioners need to take seriously.

About Demandbase

Demandbase is an account-based marketing platform built for enterprise B2B organizations. The company reached a significant milestone in 2020 when it merged with Engagio, combining capabilities to build what became one of the most comprehensive account-based marketing platforms available. Its flagship product, Demandbase One, brings together intent data, account identification, advertising, and sales intelligence in a single platform. The platform is designed to help B2B organizations identify, prioritize, and engage with the accounts most likely to generate revenue, combining firmographic, technographic, and behavioral data to surface buying signals. Labs by Demandbase is the company's dedicated research arm, built to analyze platform data at scale and publish findings that go beyond the vendor's own marketing interests. Rachel Truair, the Chief Marketing Officer at Demandbase, has played a visible and active role in bringing this research to market, which matters. When a CMO puts their name on a study that challenges conventional practice, including practices that vendor's own customers rely on, that is a signal worth noting.

What 24 billion interactions actually show

The dataset behind this research is not a survey. It draws on data from 1,452 Demandbase tenants, 429,634 advertising campaigns, 38 million marketing activities, and 9.7 million sales interactions. At that scale, the noise from individual company behavior largely cancels out. What remains are patterns that hold across thousands of go-to-market teams simultaneously.

Five findings stand out, and I want to give you my honest read on each one, not just the headlines.

Stalled accounts: stop pushing, start watching

The finding that increased outreach on stalled accounts decreases conversion is the one I resisted most. My instinct says it depends on execution: a thoughtful, well-timed message from a senior person can reopen a door. But at scale, the data shows that most organizations default to volume, not quality, when an account goes quiet. Seven "just checking in" emails do not constitute a relationship. They signal that you are not paying attention to the buyer's reality. The research is probably not telling us that persistence is wrong. It is telling us that undisciplined persistence, driven by a cadence tool rather than genuine signal, is actively harmful. The practical correction is to set stop-loss thresholds: if an account has not engaged after a defined number of touches, pause the sequence and wait for organic intent to resurface before resuming contact.

The buying group is not a new idea, but the evidence for it is now overwhelming

Organizations that align their go-to-market teams around buying groups achieve two to three times higher win rates than those focused on individual leads. A typical buying group includes 13 to 17 stakeholders, each with different priorities and levels of influence. Most marketing teams know this intellectually and still build campaigns around a single contact. The Marketing Qualified Lead persists as a metric not because it is accurate but because it is easy to measure and easy to hit. Replacing it with a qualified buying group metric requires more coordination and more uncomfortable conversations with leadership about what a successful quarter actually looks like. That coordination is exactly the point.

Advertising is not a luxury item

Accounts supported by sustained, buying-group-level advertising convert to opportunities at two to three times the rate of accounts without advertising. Companies using four advertising products report a 58.7% win rate, a 71% lift over companies using none. I have sat in enough budget reviews to know that digital advertising is one of the first items cut when pipeline looks thin. The logic seems sensible: cut brand spend, protect direct response. The data says the opposite. Sustained air cover across an active account changes the buyer's frame of reference before your sales team ever gets in the room. The CFO who has seen your value framing three times before the first call is a different audience than the one encountering you cold.

The cost of a siloed tech stack

Teams operating within more mature account-based marketing frameworks see a 22.33% median conversion rate from Marketing Qualified Account to pipeline. Less mature programs convert at 14.19%. That gap is not primarily a strategy gap. It is a data gap. When your marketing automation platform, your customer relationship management system, and your intent data cannot communicate in real time, your predictive models are working with an incomplete picture. They will generate false positives. Sales will chase the wrong accounts. The fix is not glamorous, it is governance and integration work, but the revenue impact is measurable and significant.

Timing is the variable that most campaigns ignore

An ideal customer profile is a list of companies that could buy from you. Intent data tells you which ones are looking right now. The difference between those two states determines most of your outbound conversion rate. Static territory lists processed in alphabetical order are a rounding error. A rep reaching out to an account the morning that account began actively researching a competitor is a different conversation entirely. The operational requirement is uncomfortable for most sales organizations: task lists that change daily based on signal, not weekly based on quota assignments.

What I still believe, and what I updated

Relationships still matter. A trusted voice inside an account opens doors that no algorithm identifies. But the data is telling me something I should have internalized earlier: most of what I called "relationship building" in stalled accounts was actually noise generation. Volume without signal is not persistence. It is interference.

The more important shift is this. B2B marketing has moved from a discipline where experience and intuition dominate to one where the organizations with the best data infrastructure and the discipline to act on it will consistently outperform those relying on instinct. That does not make the practitioner's judgment irrelevant. It makes the practitioner's job harder, because now judgment has to contend with evidence.

The Demandbase research is worth reading in full. The benchmarks are real, the dataset is serious, and the implications for how you build your go-to-market motion are significant whether or not you use their platform.