Analytics
The Training and Win/Loss dashboards, and what each is actually good for.
Analytics has two tabs, and they answer two different questions.
Training
Per training plan: how many reps started, how many certified, the team average with its sample size, the coaching priority, a leaderboard, score by dimension, and a Stages and steps view showing where reps stall.
This answers "is the team getting better, and at what?"
Full detail in training performance.
Win/Loss
Across your interviews: outcomes, competitors, and the recurring reasons buyers give. Reachable from Analytics → Win/Loss or Performance on the Interviews screen.
This answers "why are we losing, according to the people who decided?"
Full detail in read the results.
Agent analytics
Separate from this screen: Agents → Analytics covers run volume, success rate and cost across your agents. That's an operational dashboard — is the automation healthy and what is it costing — rather than a GTM one. See runs and debugging and credits and usage.
Reading these well
Check the sample size before the number. A team average marked n=1 is one person's morning. The dashboards
show the sample deliberately; use it.
Look for the same story in both tabs. A weak training dimension and a recurring loss theme are usually one problem seen from two ends:
| Win/Loss says | Training says | Read it as |
|---|---|---|
| Buyers cite a competitor's integration depth | Objection handling is the weakest dimension | Reps can't handle a live objection you already know is coming. Build the drill |
| Losses cluster on "no decision" | Discovery depth is weak | Qualification failure, not a closing failure |
| Won deals cite one specific proof point | Value articulation is strong | Protect that proof point when you change messaging |
Trends over snapshots. One week's movement in a team average is noise. The useful cadence is monthly for training, and per-quarter themes for win/loss.
What isn't here
Pipeline and forecast reporting isn't a fixed dashboard — it's a Live App an agent keeps current, so it matches how your team actually reports rather than a template. See the deal intelligence recipe.