Filters and date ranges
Scope every chart on a dashboard by date range, aggregation, channel, category, agent, sentiment, priority, and language.
Every dashboard shares one control bar. The date range, the aggregation, and the filters you set there apply to every chart on the page at once — and they follow you as you switch between dashboards, so you can compare the same slice across Overview, AI Performance, and your own boards without re-picking it.
Date range
Analytics opens on the last 30 days. The picker offers:
| Preset | What it covers |
|---|---|
| Today | From midnight (your local time) to now |
| Last 7 days | The rolling 7 days ending now |
| Last 30 days | The rolling 30 days ending now |
| Last 90 days | The rolling 90 days ending now |
| Custom | Any start and end date you pick on the calendar |
For a custom range, choose Custom in the picker: a two-month calendar opens in place. Click once to set the start date and once more to set the end.
A few metrics ignore the date range
Plan-usage metrics — AI conversations used, Percent of plan used, and Overage cost — always reflect your current billing period, whatever range is selected. Open conversations is a snapshot of what was open at the end of the selected range, not a total over it.
Aggregation
Time-series charts bucket their data by the aggregation you pick: Daily, Weekly, or Monthly. This changes how trend lines and bar charts group time; it doesn't change KPI totals.
Filters
Click Add filter and pick a dimension. Each active filter appears as a pill you can reopen to adjust; within a filter you can select multiple values. Use Clear to drop all filters at once, or open a pill and choose Remove filter to drop just that one.
| Filter | What it matches |
|---|---|
| Channel | The channel the conversation arrived on (chat widget, email, …) |
| Category | The conversation's categories — a conversation with several categories matches if any selected one applies |
| Agent | Conversations a selected agent took part in — replied to, closed, or was assigned. Includes your AI agent and a catch-all "Human agents" option alongside individual members |
| Sentiment | The conversation's detected sentiment |
| Priority | The conversation's priority |
| Language | The conversation's detected language |
Filters combine with and across dimensions and or within one: filtering by Channel = chat widget and Sentiment = negative shows negative-sentiment widget conversations only.
Where filters apply
The active range and filters flow through everything on the page:
- Every chart and KPI recomputes against the filtered set.
- Drill-through lists only conversations matching the active filters plus the slice you clicked.
- CSV export contains the filtered numbers — what you see is what exports.
Spam is excluded by default
Every metric excludes conversations marked as spam — except the Spam conversations metric itself, which counts only them. See How metrics are calculated.