HockeyStack raises $50M Series A to build Revenue Agents for the enterprise
HockeyStack raised $50M to expand enterprise “Revenue Agents.” See how event-based revenue data and workflow automation may reshape RevOps tooling.
HockeyStack raised $50 million in a Series A round as it rolls out what it calls “Revenue Agents” aimed at enterprise sales and RevOps teams. The funding includes participation from Bessemer Venture Partners, Y Combinator, and Uncorrelated Ventures, and brings total funding to over $50 million.
The core claim behind the launch is architectural: HockeyStack says it is moving revenue teams away from traditional CRM-style objects and fields toward an event-based data model, then using that data to drive agent-like workflows across prospecting, new business, and expansion.
Short on time?
Here’s a quick look at what’s inside:
- What HockeyStack is building with Revenue Agents
- Why event-based revenue data matters for enterprise teams
- Where HockeyStack fits versus Dreamdata and CaliberMind
- What this signals about AI-native revenue operations
- Practical takeaways for revenue and marketing ops leaders
What HockeyStack is building with Revenue Agents
HockeyStack is positioning Revenue Agents as always-on support tied to deals and accounts, with “manager views” intended for coaching and forecasting. A central component is a “Blueprint” model that ingests structured and unstructured enterprise data and is meant to update as new deals close and conditions change.
For marketers and RevOps teams, the key concept is not a generic chatbot. It is an agent layer that sits on top of a company’s revenue data and attempts to turn institutional knowledge into repeatable actions, such as prompting reps at specific moments, recommending next steps, or standardizing how teams qualify and expand accounts.
The company says it has reached over 300 customers in under two years and is targeting large enterprises, citing adoption by “Fortune 100 revenue teams” and named customers including 8x8, DataRobot, AppsFlyer, Rakuten, RingCentral, LaunchDarkly, and Mastercard.
ContentGrip
Why event-based revenue data matters for enterprise teams
Most revenue workflows still revolve around CRM records and manually maintained fields, plus a patchwork of enrichment, attribution, and BI layers. The practical limitation is that many meaningful buyer and seller signals are inherently event-like: web activity, product usage, conversation outcomes, buying committee changes, and expansion triggers. When those signals are flattened into static fields, teams can lose sequence, timing, and context.
An event-based model can make it easier to answer operational questions that matter for pipeline quality, such as: which sequence of activities tends to precede late-stage conversion in a specific segment, what actions typically correlate with expansion, or how quickly deal momentum decays after certain stalls. If HockeyStack can reliably connect those event streams to “agents” that recommend or execute actions, it is essentially betting that the system of record evolves from data entry to decision support.
That framing also explains why this round is tied to “enterprise” messaging: large organizations have more fragmented tools and more variance across business units. Systems that can adapt to different revenue motions without heavy reconfiguration tend to have an advantage in complex environments.
Where HockeyStack fits versus Dreamdata and CaliberMind
HockeyStack competes in a crowded revenue technology landscape where many vendors aim to unify revenue data and make it actionable. Companies like Dreamdata and CaliberMind are commonly associated with B2B revenue attribution and multi-touch reporting, helping teams connect marketing touchpoints to pipeline and revenue outcomes.
HockeyStack’s differentiation, based on its positioning, is less about attribution dashboards and more about an agent-driven workflow layer tied to an event-based foundation. In other words, the competitive question is whether teams want primarily measurement (attribution and reporting) or orchestration (recommended actions and automated execution) built on top of unified signals. In practice, enterprises often need both, which increases pressure on vendors to expand beyond their initial wedge and converge into broader “revenue platforms.”
The intensity of the category also means procurement scrutiny will be high. Buyers will look for clear proof that an event-based architecture improves forecast accuracy, sales productivity, or expansion conversion, not just that it produces more data.
What this signals about AI-native revenue operations
Two trends are converging here. First, AI-native SaaS is pushing vendors to bundle analysis and execution into the same interface, since “insight without action” is increasingly seen as wasted cycle time. Second, revenue tech convergence is accelerating as attribution, data unification, enablement, and workflow automation blur into one operating layer for go-to-market teams.
Revenue agents are one expression of that convergence: a vendor tries to own the data layer, the decision layer, and parts of the execution layer. For marketing leaders, the implication is that boundaries between marketing ops, RevOps, and sales ops will keep softening as tooling shifts from channel-specific systems to account and revenue systems.
Practical takeaways for revenue and marketing ops leaders
If you are evaluating agent-led revenue tooling, a few practical checks can reduce risk:
Start with narrow, measurable use cases. For example, expansion trigger identification for a single product line, or consistent follow-up on hand-raisers in one segment. Tie success to metrics like speed-to-lead, meeting-to-opportunity rate, stage progression time, or renewal and expansion outcomes.
Audit data readiness. An “agent” is only as good as the event instrumentation and identity resolution beneath it. Confirm what sources can be ingested (product events, web, email, meetings, call transcripts) and how conflicts are resolved across systems.
Clarify what is automated versus suggested. Enterprises often prefer recommendation-first deployment before granting execution permissions. Map what actions the system will take, what approvals are needed, and how exceptions are handled.
Demand governance and change management. If models “update continuously,” teams need visibility into what changed and why, especially when outputs affect forecasting, routing, or account prioritization.
