The B2B intent data market sold a story for a decade: anonymous research activity reveals who’s about to buy. Most reps who used the data quietly know it doesn’t work the way it was promised. The signals that actually predict a buying conversation aren’t anonymous and aren’t statistical. They’re named events with dates and dollar amounts attached.
Any named individuals shown in this post are real people surfaced through Lusha’s signals layer. Last names abbreviated for privacy. Full records — including emails and direct dials — are returned inside the user’s Claude session.
A few weeks ago a sales leader showed me her intent data dashboard. The dashboard said 47 accounts in her territory were “showing high intent” that week. She’d been running this dashboard for nine months. She’d opened maybe four conversations from it. None had closed. She kept the subscription because nobody had built her anything better, but she’d stopped checking the dashboard daily because the signals didn’t translate into anything she could act on.
This is the open secret of B2B intent data. The signals are statistical — derived from anonymous web-tracking pixels, surge models, and behavioral aggregates. The output is “this account is researching topics related to your category.” The buyer behind the activity is unnamed. The intent itself is inferred. The timing is approximate. The rep gets a list of accounts and a confidence score, with nothing concrete to lead with.
The real buying signals — the ones that predict an actual conversation — look completely different. They’re named. They’re time-stamped. They name specific people doing specific things at specific companies on specific dates. And they’re public, indexed, and verifiable. They’ve always existed. The market just spent a decade selling anonymized inference instead.
What named signals look like
Pulled from the last six months of indexed B2B activity, here are real signal events the buying signals gallery prompts surface for active accounts:
Snowflake. New Chief Revenue Officer Jonathan B. started March 31, 2026 — internal hire, replaced Mike G. who left for personal reasons. Brand-new Chief Security & Trust Officer Mayank U. started April 1, 2026 — a role that didn’t exist at the company before. Three acquisitions in six months — TensorStax (agentic AI for data engineering), Select Star (data catalog), Observe (AI observability) — all in the AI data infrastructure space. Six product launches including the Cortex Code Agent SDK. Operations hiring +60% against baseline, Marketing +57%, Sales +29%. Web traffic +279% on February 12, 2026.
MongoDB. Ryan M. appointed Chief Revenue Officer effective April 27, 2026 — hired “to support next phase of growth,” per the company’s announcement. CEO Dev I. stepped down November 10, 2025, replaced by CJ D. Regional Sales Director departed in May. Three leadership changes in the GTM function inside six months.
Verkada. CapitalG-led $5.8B valuation funding round, December 2025. New VP of EMEA Mark C. joined February 2026. New Head of Middle East with Dubai office opening, January 2026. Sales hiring +37%, Engineering +25%.
Together AI. $1B funding round in negotiation, valuation eyed at $7.5B, per March 6, 2026 coverage. NVIDIA-powered infrastructure positioning. Sales hiring +23% against baseline.
Notion. Sales hiring +204% against a 16.7-job baseline — 51 new sales jobs in four weeks, the highest-intensity hiring surge in the territory.
Every one of these is a public, time-stamped, named event with a source article URL. None of them is “Notion is researching topics related to revenue operations.” All of them are useful to a rep. The difference matters.
Why named signals convert and anonymous intent doesn’t
Three reasons.
01. A named event tells the rep what to lead with. When the rep knows Snowflake has a new CRO who started six weeks ago, the opener writes itself: “Saw the CRO appointment in March — given Jonathan’s mandate to scope the next phase of GTM, wanted to share where the conversation left off.” When the rep knows Together AI is in active negotiation on a $1B round, the opener writes itself: “The funding round in negotiation changes the conversation about budget appetite. Worth picking up where we left off?” When the rep has anonymous intent data showing “high research activity in the past 7 days,” the opener doesn’t write itself — the rep has to manufacture context the data doesn’t provide.
02. A named event creates conversation legitimacy. A prospect who receives an email referencing a public event their company actually did reads the email as informed. A prospect who receives an email that says “I noticed your team has been researching this category” reads the email as surveillance — because the rep has demonstrated knowledge the prospect didn’t authorize them to have. The first lands. The second triggers either no response or a “how did you get this information” question that ends the conversation immediately.
03. A named event is structurally tied to buying authority. A new CRO has a mandate window. A new CSTO has a procurement gate forming around their role. A fresh funding round unlocks specific budget allocation 60-90 days out. An M&A event triggers vendor rationalization on a roughly 12-month timeline. Each named signal carries an inherent timeline that maps to the buyer’s actual decision-making calendar. Anonymous intent has none of this — the rep doesn’t know who at the account is researching, why, with what authority, on what budget cycle, or against what alternatives.
What named signals can do that intent data can’t
The buying signals workflows in the prompt gallery answer questions intent data structurally can’t.
“Which of my target accounts has a new CRO in their first 90 days right now?” The signal layer returns the named executives who took the role recently, with start dates, source URLs, and the verified buying group below them. The rep walks into the conversation knowing exactly which mandate window is open. (See this workflow →)
“This stuck deal went quiet two months ago. What signals have fired at the prospect since then?” The signal layer scans the prospect for events fired after the last meaningful touch — new leadership, M&A, product launches, hiring surges. Each one is mapped to a specific re-engagement angle the rep can use as the email opener. Replaces the dead-letter “just checking in” with a real reason to re-open. (See this workflow →)
“What’s happening at Snowflake right now? I have a discovery call in two days.” The single-account signal brief returns the complete 6-month picture — leadership moves, strategic acquisitions, product launches, hiring intensity, market signals — organized into five categories with talk tracks tied to each. The kind of brief that takes a rep three hours of LinkedIn scrolling to assemble manually. (See this workflow →)
“Which of my customers is showing both expansion AND risk signals at the same time?” The dual-state scan surfaces mixed-state accounts — Snowflake firing both expansion-positive signals (new CSTO role, three AI acquisitions) AND renewal-risk signals (CRO departure, layoffs, class-action lawsuits) simultaneously. That’s the real CSM intelligence — not a single score, but two simultaneous reads on the same account. (See this workflow →)
None of these workflows would be possible with anonymous intent data alone. The questions reps actually ask require named entities, dates, and verifiable sources. The signal layer provides them. The AI agent organizes them. The rep acts on them.
What’s actually changed in the category
For most of the last decade, B2B intent data was sold as the buying-signal solution because nothing else was operationally available. Web-tracking pixels were the only scalable signal source. Surge models were the only way to turn anonymous activity into a sortable list. Named-event tracking existed — funding databases, news monitoring, executive-move trackers — but the data lived in separate tools, didn’t refresh continuously, and wasn’t queryable from inside the rep’s actual workflow.
The 2026 unlock is that named-event data is now both continuously refreshed and queryable from inside the AI agent. The signal layer that Lusha exposes through the MCP connector indexes funding rounds, executive moves, M&A events, product launches, hiring surges by department, and headcount shifts — all time-stamped, all source-attributed, all tied to verified contact records at the named entities. The same data layer that returns the verified contact list also returns the signal context for each contact’s company.
When a rep asks Claude “find me CROs at SaaS companies that raised funding in the last 90 days”, the workflow runs against the same data infrastructure that returns the contact list. The signal isn’t a separate intent feed bolted onto the prospecting tool. The signal is part of the same verified data layer the rest of the workflow runs on.
That integration matters because it removes the structural reason most sales orgs split intent data and contact data into separate tools — they came from separate vendors with separate data models. When the same layer returns both, the workflow becomes single-step. The rep asks the question. The answer comes back complete. (See the buying signals gallery →)
What to try next
The fastest test of the argument is to run one named-signal workflow against an account in your active pipeline. Pick the account. Ask Claude to surface the events fired in the last 90 days — leadership moves, funding, M&A, hiring surges, product launches. Compare what comes back to what your current intent dashboard shows for the same account. The difference is usually obvious.
If the test lands, the buying signals gallery covers the eight specific workflows in depth — target account signal scans, contact role monitoring, single-account briefs, weekly digests, stacked-signal account discovery, closed-lost re-engagement, customer expansion signals, and renewal-risk scans. Each prompt page includes a live demo of the workflow running against real Lusha data. (See all buying signal prompts →)
For teams running customer health monitoring specifically, the Customer Health to Action Skill packages the expansion-and-renewal-risk workflows into a single Claude Project with Gmail draft generation built in. The same scan that surfaces a customer’s expansion-ready triggers also surfaces the renewal risks — and drafts the role-specific outreach for both. (See the Skill →)
The argument isn’t that B2B intent data has no use. The argument is that anonymous research activity has been over-sold as a buying signal for a decade, and the real signals — the ones reps can actually open conversations with — have always been named events on indexed timelines. The category just needed a workflow surface that could query them in plain English and a data layer that kept them current.
Both exist now. The signal layer is real, the agent is real, and the workflow runs in one chat.
