Buyer signals only turn into pipeline when a seller knows who to call and what to say.
Datadog‘s Kevin Driscoll, who leads global ABM and demand generation, joined TechInformed’s Madelaine Oppert, SVP of Global Marketing on September 24 to show how his team makes that happen.
Key takeaways
- Person-based ABM adds a second layer to your account list: the individuals most likely to take a meeting.
- Collect every signal, but score fewer than 10, weighted by how well each predicts readiness to buy.
- Most signals stay live for at least a month, which leaves time to plan outreach.
- Give sales context and a suggested message with every signal, and trust a good rep’s instinct when it clashes with the data.
- Meetings give faster feedback than pipeline, so track second-meeting rates against a no-signal baseline.
Plenty of buyer signals, little direction
Most B2B teams have more buyer signals than they can use, from intent data and account scores to job changes and website visits.
The trouble starts at handoff, when sales hear an account is warm and still has no idea who to call or what to say.
This live discussion covered which buyer signals predict a meeting, how to reach the right person inside an account, and how to tell whether those meetings turn into pipeline.
Person-based ABM starts once the account list is set
“Figure out which people inside of those accounts are going to be most impactful and how do we activate them individually.”
Madelaine opened by asking how the model changes when prioritization moves from account to person, and Kevin called it the biggest mind shift from traditional ABM.
Every account holds several buying groups with different levels of warmth toward your brand, so the work becomes finding the people most willing to talk right now.
Madelaine added that the person creating friction, such as someone in legal, finance, or operations, may sit outside your usual audience. To find them, Kevin reverse engineers which buyer signals fired as recent customers bought.
A signal earns its weight by predicting a meeting
“How critical is this signal to telling you whether or not they’re ready to buy right now?”
Kevin prefers that question to the phrase “weak signals.” One white paper download says little, yet seven of them start to look a lot like a trial request.
When an audience member asked whether there’s such a thing as too many buyer signals, Kevin admitted he’s a bit of a hoarder and advised collecting whatever you can, since data sources and AI keep improving. However, the model you score against should stay in the single digits. There’s a difference between collecting signals and actioning them, and that the signals you can’t use right now, may be able to be used in two years’ time, so hold onto them!
Your go-to-market motion sets the thresholds, so sellers at Datadog “jump on just about any breadcrumb” because an engineer starting a trial carries real weight in a bottoms-up culture.
Sales needs context with every signal
“These signals should never be provided to sales without context or without multiple threads at once.”
A common struggle for sales teams everywhere–right? Firing funding announcements and lead alerts at sellers produces the complaint Kevin hears most, which is knowing an account is warm with no idea what to do next.
His answer is to layer signals into a “constant account heartbeat” built around individual people so sellers know why this person and what message to lead with.
Timing outreach without crossing the creepy line
“What you want it to look like is a happy coincidence.”
Kevin splits buyer signals into observable and unobservable. A public LinkedIn post is fair to mention in a message, while anonymous web traffic and bidstream intent should stay out of it. Instead, the more natural the signal appears, at the right time, the better.
Outreach informed by those can arrive as news about a feature tied to something the person already cares about. In Kevin’s words, “You just need to deliver the relevance.”
Buyer signals stay useful longer than LinkedIn suggests
“It’s a lot longer than the LinkedIn gurus will tell you.”
Kevin’s team finds that leads, form fills, job changes, and funding announcements stay live for at least a month, which leaves time to plan a strategy before anyone reaches out.
Website traffic has gotten noisier with AI and deserves more caution.
He also doesn’t treat AEO or GEO as a signal source today because AI platforms remain “a real black box.”
Want to see how AI is changing the way buyers research? Watch our on demand session, Your buyers are using AI. Is your demand gen strategy ready?, featuring Elena Simkina from Acronis.
When the data and the rep disagree
“If a rep’s intuition, especially good reps, is telling you that the data is wrong, then they’re probably right.”
Kevin put it bluntly: “Data is going to lie to you.”
At a previous company, an account kept researching documentation, APIs, and integrations yet never wanted the product. The buyer finally explained six months later that their CEO preferred another vendor and the team had only used Kevin’s company to fill out an RFP.
The rep had sensed it all along, so Kevin’s advice is to let reps tell you what matters and use your data to back them up.
Meetings only matter if you test them
“Meetings are not pipeline. Meetings are not revenue.”
Meetings give faster feedback than pipeline, yet Kevin named that very assumption as the one deserving the most scrutiny.
He values a quick, clear no because it frees rep time, and he watches second-meeting rate closely, comparing ABM-sourced meetings with everything else.
After a first meeting, renewed website activity or new inbound leads point to progress, while silence on both almost always means a no.
If meetings rise and qualified pipeline doesn’t, he checks the stalled meetings and deal size, and he tests the signals by asking “what happened if you had no signal?”
Three things to try this week
Madelaine asked for a checklist, and Kevin offered three steps:
Analyze your closed-won deals
Run your last 10, 20, or 30 closed-won deals and their meeting, sales, and ops notes through an LLM to find a new signal or a better channel strategy.
Build your own buyer signal
Build one signal tailored to your business, with AI or a partner. Datadog monitors outages, since reliability trouble is usually when prospects need its platform.
Rethink your channel mix
Check your channel mix against how your buyers buy, because LinkedIn only works when your people are on it.
FAQs
What is person-based ABM?
It keeps the target account list of traditional ABM and adds a step that identifies which individuals inside each account are most likely to take a meeting.
How many buyer signals should go into a scoring model?
Kevin Driscoll from Datadog recommends a single-digit number, with the exact count depending on your business and buyer.
How long does a buyer signal stay actionable?
It varies, although Kevin Driscoll’s team finds that leads, form fills, job changes, and funding announcements stay live for at least a month.
How can you use intent data without sounding creepy?
Mention only observable signals, like a public LinkedIn post, and let outreach informed by anonymous web traffic arrive as a relevant, well-timed message.
Are meetings a reliable measure of ABM performance?
They give faster feedback than pipeline but remain a leading indicator, so track second-meeting rate and compare results against a no-signal baseline.
Watch the full session
Watch the full discussion on demand for a complete breakdown of Kevin’s insights and advice, and follow Kevin Driscoll on LinkedIn for more of his thinking.
