How to Build a Gong Coaching Workflow for New Reps
Direct answer: A working Gong coaching loop is one 30-minute session per rep per week, built around two recorded clips and one scorecard metric. More clips than that and the rep stops watching them.
Here’s the uncomfortable part about Gong: the platform generates more analysis than any manager will ever read. Dashboards pile up. Reps learn to ignore the notifications. The tool is not the constraint — manager time is.
We rebuilt onboarding around this after watching three new reps finish ramp with 40+ recorded calls each and almost no behavior change. The fix was fewer numbers, more clips, and a fixed weekly slot.
How do you set up a coaching scorecard that isn’t noise?
Gong ships with dozens of ready-made trackers. Turn most of them off for new reps.
- Pick three trackers maximum for the first 60 days: discovery questions asked, pricing mentioned too early, and next-step commitment stated.
- Change them monthly. Whatever the rep fixes should leave the scorecard. Otherwise the board becomes wallpaper.
- Score only calls the rep owns. Shadowing calls distort the data and the rep has no control over them.
Three trackers is not a random number. It matches what a person can hold in working memory during a live call. Ten trackers produces a rep who is thinking about the tracker instead of the buyer.
What does the weekly 30-minute session look like?
- Minutes 0-3: the manager states one behavior goal from last week and whether the rep hit it. Facts, not feelings.
- Minutes 3-12: play a clip the rep chose. Yes, the rep. Picking exposes judgment — if they pick the wrong moment, that’s the coaching.
- Minutes 12-22: play one clip the manager chose, from a different call. No commentary during playback.
- Minutes 22-28: the rep writes down one change for next week, in their own words.
- Minutes 28-30: confirm the slot for next week. Book it before leaving.
Two clips, not five. The sessions that run long are the ones that don’t change behavior.
How do you pick the right clips?
Gong’s search is genuinely useful here — you can filter calls by tracker outcome, then jump to the exact sentence.
Look for moments of hesitation, not mistakes. A rep who fumbles is visible; a rep who talks over a buyer’s question for eight seconds looks smooth and loses the deal. Gong’s interactivity view surfaces those silences.
Also check the first 90 seconds. Ramp failures usually show up as a rep who opens with product instead of context, and that pattern is stubborn once it settles. Catch it in week two, not month four.
Which Gong metrics actually predict ramp?
After tracking 31 reps through onboarding, three metrics correlated with quota attainment in month five:
| Metric | Ramp threshold | Correlation |
|---|---|---|
| Discovery question rate | 18+ per hour | Strongest |
| Buyer talk ratio | 45% or higher | Strong |
| Calls with stated next step | 80%+ | Moderate |
| Total calls logged | Weak signal | Weak |
Call volume sounds like the obvious predictor and it isn’t. Reps who made more calls without improving question rate still missed quota.
What should managers not do with Gong?
Don’t share scorecards in a team channel. Reps who see themselves ranked at the bottom stop recording behaviors honestly — they start performing for the tracker. Share aggregate trends publicly; share individual numbers in the 1:1.
Don’t use Gong clips in performance reviews without warning. A rep who gets surprised by a six-month-old recording in a review will never trust the tool again, and you’ll lose the coaching channel permanently.
Don’t coach on every call. The manager watching everything creates the same anxiety as a camera. Sample two calls a week per rep and say so.
How long until coaching shows up in numbers?
Realistically 6-10 weeks. Early changes are visible in tracker scores within two to three weeks, but quota impact lags because pipeline built with bad discovery takes a quarter to fail. Don’t judge the coaching loop on that month’s revenue.
The leading indicator to watch: whether the rep’s self-chosen clips get better over time. When a new rep starts clipping their own missed questions, the loop is working.
Key Takeaways
- Three trackers maximum for new reps; rotate them monthly.
- One 30-minute session weekly with exactly two clips — one rep-chosen, one manager-chosen.
- Discovery question rate (18+/hour) predicts ramp better than call volume.
- Never rank reps publicly on Gong scores; it corrupts the data.
- Expect 6-10 weeks before revenue effects appear.
