The Analytics dashboard — what every metric means
Total interviews, completion rate, average score, average duration, and the funnel of pending → completed → expired → cancelled. The full read of every number on the Analytics page.
Key takeaways
- Filters at the top: Job Position, Start Date, End Date — or quick presets (7/30/90 days, This Year).
- Key Metrics row gives you the overall health: total, completion %, average score, and average duration.
- Status funnel below shows Pending, Completed, Expired, Cancelled — read these together to spot drop-off.
- Tweak the Start/End filter aggressively — the rolling 30-day view tells a very different story from "this year."
Before you start
The Analytics page is your aggregate view of interview activity — useful for spotting trends, comparing roles, and reporting up to leadership. This article walks through every number on the page and how to read it.
Filters (top of page)
Three filters and four shortcut buttons:
- Job Position — defaults to All jobs. Pick a specific job to drill into one role's hiring funnel.
- Start Date / End Date — manual range. Use this for arbitrary windows (e.g. a single hiring sprint).
- Quick presets — Last 7 / 30 / 90 Days, This Year. The default is Last 90 Days for most accounts.
TIP
Resist the urge to look at *This Year* by default. A rolling 30- or 90-day view is more honest about current performance — *This Year* averages your current process with whatever you were doing in January, which may have been very different.
Key Metrics row
Total Interviews
Count of all interview invitations in the filter window (NOT just completed). The big number is the total; the small subtitle splits it: X completed · Y% completion rate.
If your Total Interviews is climbing but Completion Rate is dropping, candidates are accepting invites but failing to finish — common causes: deadline too tight, technical issues during pre-flight, candidates losing interest.
Completion Rate
Percentage of invitations that ended in a completed interview (transcript + evaluation generated). The subtitle shows how many expired and failed.
Benchmarks:
- >85% — healthy. Most candidates who accept actually finish.
- 70–85% — normal range, especially during seasonal hiring rushes.
- <70% — investigate. Are invites timing out? Is the candidate pool too senior to invest 30 minutes? Are technical issues rampant?
Average Score
Mean of overall-recommendation scores across all completed interviews in the window. The subtitle shows the range (min–max).
A wide range (e.g. 7–88) is actually healthier than a tight range (e.g. 60–70) — wide ranges mean the AI is discriminating between strong and weak candidates. A tight range can mean either the candidate pool is homogeneous (good) or the evaluation is poorly calibrated to your role (bad — fix the JD or the role weights).
Average Duration
Actual time spent in the interview room across all completed interviews. The subtitle says Per interview.
If this is consistently 50–60% of your configured slot, the AI is wrapping up early — usually because it covered the planned questions or your role doesn't have enough material to fill the slot. Switch to a shorter duration to save credits.
Status funnel (second row)
Pending, Completed, Expired, Cancelled. These should add up to roughly your Total Interviews.
- Pending — invited but not yet started. Healthy at low single digits in a steady state; large numbers mean your invitations are sitting in inboxes unactioned.
- Completed — finished and evaluated. The desired terminal state.
- Expired — link aged out before the candidate started. Default expiry is 7 days; adjust per-job in job settings.
- Cancelled — manually revoked. Track this if you suspect your team is over-inviting and then realising mistakes.
What to do with this view
Weekly review
Once a week, open Analytics with the Last 7 Days filter. Scan the four key metrics. Anything that swung 20%+ from last week deserves five minutes of investigation.
Per-job comparison
Drop the Job Position filter to a specific role. Compare completion rate + average score across roles. Roles with consistently low completion rates often have JD issues (too aggressive, unclear, too long); roles with low average scores often have a candidate-sourcing problem.
Reporting up
For leadership reports, screenshot the 90-day or quarterly view. The Total Interviews + Completion Rate + Average Score combo is the cleanest health snapshot.
Next: how to slice these metrics by filter combinations to answer specific questions.
Frequently asked questions
Why is my Average Score so low?
Average Score is the mean of overall-recommendation scores across all completed interviews in the filter window. Two common reasons it looks low: (1) you're including failed interviews where the AI couldn't generate a full evaluation (these score near zero), or (2) your candidate pool genuinely scored low and you should look at the JD/screening filter upstream.
Average Duration is shorter than my configured interview slot — bug?
No. Average Duration is the *actual* time spent, not the *scheduled* time. If your jobs are configured for 45 min but candidates often finish in 30, the AI is wrapping up early because it covered the planned questions. That's fine for short interviews; if it's happening on senior roles where you wanted depth, check your interview-plan settings.
What counts as Expired vs Cancelled?
Expired = invitation link aged out without the candidate starting (default 7 days). Cancelled = you (or a teammate) manually revoked the invitation. Use Expired to gauge whether your invitation window is too tight; use Cancelled to track recruiter behaviour.