Agents Every call scored, not just the ten someone had time to hear
It is 9:05 a.m. on a Tuesday and the office manager at a fourteen-truck HVAC and plumbing shop is staring at last week’s call report: hundreds of calls in the system, and coaching notes built from the ten recordings someone had time to hear.
Source: Netic, "How Paschal doubled its revenue without doubling its call center" — https://www.netic.ai/customer-stories/paschal — the story describes a shop fielding 4,000 to 5,000 calls a week outside peak season, where managers could hear maybe ten calls per rep each week, and an AI that now scores every call so coaching moves from anecdotes to trends.
The problem
Call coaching in most shops is a sampling exercise. The manager pulls a handful of recordings when there is a quiet hour, hears the best rep on a good day and the newest rep on a bad one, and coaches from that slice. The other calls — nights, weekends, the heat-wave rush — never get heard at all.
So the same booking misses repeat invisibly: a price quoted before the problem is understood, no email captured for the estimate, no maintenance plan offered on an aging system, no second attempt when the first slot does not fit. Each one is small. Across every call, they are the difference between a full board and a board with holes.
How the agent workflow works
- The shop writes its own checklist first. The owner and office manager define a good call in their words: greet and identify, diagnose before pricing, capture name, address and email, offer the maintenance plan where it fits, confirm the slot and arrival window.
- Every recorded call gets scored against it. The agent works through the full call log — inbound, outbound, after-hours — and marks each item done, missed or not applicable, with the moment in the recording where the evidence sits.
- Misses are clustered, not just counted. The agent groups the patterns: price quoted too early on replacement calls, email missing on estimate bookings, plan offers skipped on Friday afternoons — by rep, call type and time of day.
- Each rep gets a one-page weekly brief. A short coaching sheet per person: two things done consistently well, the one miss that showed up most, and two or three actual calls worth listening to together. The agent picks the examples; the manager does the coaching.
- The checklist itself gets reviewed monthly. If every rep misses the same item, the brief says so plainly — that is usually a checklist, training or pricing problem, not a people problem — and the owner decides what changes.
What you need to set it up
- Call recording and transcription already running on your phone system, with customers and staff aware calls are recorded under your state’s rules.
- A booking checklist written by the shop — five to eight items a manager can defend — not a generic scorecard copied from a vendor.
- Access for the scoring tool to the recordings and the booking system, so a scored call can be tied to whether the job actually booked.
- A manager with a standing half hour a week to read the briefs and coach from them; the brief does not coach anyone by itself.
- Honest costing: the source publishes no price for this scoring. Expect it inside a call-platform or AI front-office subscription, plus the manager’s review time each week.
Where it breaks
A score is not a judgment about the call. The agent can hear that no email was captured; it cannot always hear why — the customer was driving, the system was down, the caller was a landlord who books differently. Scores handed down without that context will lose the room fast, and reps will stop trusting every number, including the fair ones.
Transcripts stumble on real calls. Cross-talk, speakerphones, accents, trade terms and part numbers all degrade transcription, and an item can be marked missed because the words were never captured cleanly. Spot-check flagged calls before any score is used in a review that matters.
Full coverage can become surveillance. Scoring every call to coach is one thing; using it to rank and punish is another, and staff will feel the difference. Keep the briefs coaching-first, share each rep’s scores with them, and never let a score alone drive discipline.
The bottom line
Run this when call volume has outgrown what a manager can sample honestly. Let the agent score every call against your own checklist and cluster the misses, and keep the human doing the part that changes behaviour — sitting down, listening together, and coaching.