The problem
The group runs a high-volume remote sales floor fed by paid social and search leads. How quickly a new lead gets its first call is the number a sales leader watches first.
The CRM had a field for it, filled in automatically. Nobody had checked that field against the phone system, and no response-time target had been agreed.
What we built
- Measurement tooling that takes lead-creation time from the CRM and the first outbound call from the telephony platform's call log (1-second precision), matched on the last nine digits of the phone number.
- Three views of the same number: wall-clock, leads that arrived while the sales floor was dialling, and adjusted for business hours.
- Segmenting, so bulk-imported contacts do not distort the metric.
- A re-runnable report with the method written down, so the number can be reproduced rather than trusted.
Result
- In a 7-day window: 124 leads created, 112 dialled (96.6%).
- Median time to first call: 11 h 01 m wall-clock; 1 h 24 m for leads that arrived while the floor was dialling (n = 37); 3 h 16 m adjusted for business hours.
- The CRM's auto-populated first-call field matched the true first dial within 30 seconds in only 39% of 930 leads. In 61% it disagreed by more than that.
Scope and limits
- This is measurement. We have no follow-up baseline and do not claim that response time improved afterwards.
- The first-call field was written by an integration we had built earlier. The finding is about that field, not about how the client's team enters data.
- Per-rep breakdowns exist in the output and are deliberately not published.