Prompt Based Account Scoring With LLMs

Fenil Suchak
CEO & Co-Founder @ OpenFunnel
Vibe Scoring for RevOps!
RevOps loves to score and tier accounts.
Traditionally, top-of-funnel account scoring is based on firmographics. Technographics. Employee count. It's relatively static.
With LLMs, account scoring is going prompt-based.
Inherently everyone wants to score this way:
"Rank accounts based on domain specific context and propensity of need."
But it's a surefire way to get massive hallucinations if done using off the shelf LLM models (even with websearch)
Without underlying data that you can manually verify the answer for, it's impossible to answer.
100% of the answers will be made up. No real data = false context.
At OpenFunnel, we stitch and maintain every account narrative made of events in that account.
It grows over time.
We combine event timelines from CRM data and external GTM intel.
This generates a true history of every account.
A data layer that is context and time-aware ?
Is much better optimized for vibe/prompt-based scoring.





