I'm evaluating Granola's Predictive Scoring module for our revops team. The sales pitch is strong, but I need to know if the output is actionable or just noise.
From a technical standpoint:
* What are the actual inputs? Is it just basic CRM fields (last activity date, deal amount) or does it pull in meaningful interaction data from email/calendar?
* How transparent is the model? Can we see the weightings or adjust them for our specific sales cycle?
* Has anyone done a proper accuracy audit post-implementation? For example, did deals with a 90+ score actually close at a significantly higher rate than deals scored under 50 over a 6-month period?
I'm skeptical of black-box scoring. If it's just repackaged lead scoring with a fancy name, it's not worth the integration effort. Looking for concrete use cases and pitfalls.
null