Analyticsintermediate

FAQ — Analytics numbers don't match what I see in the inbox

The two views answer different questions. The inbox shows live state at this moment; Analytics shows historical state at conversation close. Both are correct, just measuring different things. Plus four other reasons numbers might disagree.

May 10, 2026

FAQ — Analytics numbers don’t match what I see in the inbox

Common when people first explore Analytics. The numbers don’t match because the two views answer different questions.

The fundamental difference

Shows — Live state right now — Historical state at conversation close
Updates — In real time — After conversations close
Counts — What exists today — What happened during a date range
SLA badge — Current trajectory — Final outcome (Met / Breached)

A conversation currently At-risk on its SLA may eventually close as Met (no breach). The inbox shows it as At-risk now. Analytics counts it as Met after it resolves. Both are correct.

Five other reasons numbers don’t match

1. Date range scoping

The inbox doesn’t have an explicit date range — it shows current open conversations regardless of when they were created. Analytics is always scoped to a specific date range. So:

Analytics says “30 conversations this week” → conversations created this week
Inbox shows “120 open conversations” → all currently-active conversations regardless of creation date

These are different numbers measuring different things.

2. Filter mismatch

Analytics filters (channel, team, country, brand) may not be set the same as your inbox filter. A “20 conversations on email today” Analytics number filtered to your team won’t match an inbox total that includes other teams’ email.

Always cross-check filter state when reconciling.

3. Closed-then-reopened conversations

A conversation that was Done last week and reopened this week:

Analytics counts it once in last week’s “resolved” total
Analytics counts it again in this week’s “active” total
Inbox shows it as currently Active

The two views aren’t double-counting; they’re measuring at different points.

4. SLA Met vs At-risk in real time

A specific failure mode in SLA dashboards:

Inbox: 5 conversations currently at-risk
Analytics: Last week’s breach rate was 1%

The 5 at-risk conversations might all eventually resolve as Met. They show At-risk in the inbox because that’s their current state. Analytics will count them as Met when they close — the breach rate stays at 1%.

If breach rate matters to you, watch it in Analytics over time. The live At-risk count is leading-indicator anxiety, not historical truth.

5. Auto-archived or auto-cleaned conversations

Conversations that were auto-archived (e.g., the “auto-close inactive” automation) may be excluded from some inbox views by default. Analytics still counts them. So an inbox showing “0 archived conversations this week” while Analytics shows 200 archived this week isn’t a bug — it’s the inbox view filter.

How to reconcile

Two-step diagnostic:

Apply identical filters to both surfaces. Same date range, same channel, same team. If they still don’t match, proceed.
Check the metric definition. Resolved conversations during the period vs. currently-active conversations are fundamentally different things. Make sure you’re comparing apples to apples.

If you’ve matched filters and confirmed you’re comparing the same metric and they still differ by more than ~2-3% (which can be expected drift due to the timing of when Analytics’ data warehouse refreshes), there might be an actual data issue. File a support ticket with both numbers, the filter setup, and the date range.

When in doubt, trust Analytics for historical questions

For “how did we do last week?” type questions, Analytics is the source of truth. It has the historical reconciliation logic. The inbox is great for live monitoring; not as great for “how many conversations did we resolve last Tuesday” — that needs the historical view.