The situation
A Series B marketing analytics platform had a strong SDR team stuck at a fraction of its capacity. Reps spent mornings building call lists by hand, cross-referencing the CRM against spreadsheets, and dialing 80 to 120 numbers a day. Connect rates were fine. Volume was the bottleneck, and the volume problem was actually a data problem.
What we built
Project Nova was an end-to-end rebuild of the calling motion:
- Data layer first. We resolved and deduplicated the account universe, enriched every contact with direct-dial numbers through a waterfall of providers, and validated the numbers before they ever reached a rep.
- Queue engine. A prioritized dialing queue, refreshed nightly, scored by ICP fit and signal recency. Reps opened their screen at 9am to a ready list. No building, no guessing.
- Parallel dialing infrastructure wired into the queue, with dispositions flowing straight back into the CRM. Every call outcome enriched the next day's queue.
- Meeting-first feedback loop. Booked meetings fed a scoring model that re-weighted the queue weekly, so the system got sharper with every conversation.
The result
Every rep on the team moved from ~100 dials to 1,500+ dials per day. Each SDR booked two or more meetings daily. The system compounded: more conversations produced better scoring data, which produced better queues.
Why it held up
The dialer was never the point. The data layer underneath it was. Because every number was validated and every account identity-resolved before dialing, the team scaled volume 10x without burning the domain, the phone numbers, or the team.