Summary
A cohort chart groups your records by when they started (for example, the month a prospect was created) and shows how a measure evolves in each period after that — the classic way to read retention, repeat engagement, or ramp-up. In Alta you build one in the Data Explorer by picking the Cohort visualization.
Who this is for
Anyone analyzing behavior over time: retention by signup month, repeat replies by campaign start week, revenue ramp by deal-close quarter.
Before you start
You need a data source (a metric or model) that has two date fields — one for when the record joined the cohort (e.g. created date) and one for when the measured activity happened — plus a measure to track.
To save the chart to a dashboard, you need permission to add widgets to that dashboard.
Step 1 — Open the Data Explorer and pick Cohort
In the left sidebar, open Explore and click Data Explorer.
Choose your data source, then select Cohort in the Visualization dropdown. The editor below changes to the cohort settings.
Step 2 — Define the Cohort Group
The Cohort Group section defines the rows of the chart:
Cohort size — a measure that counts how many records started in each period (e.g. prospects created). This is the denominator when values are shown as percentages.
Cohort date — the date field that assigns each record to a cohort. Records sharing the same bucket (same month, week, etc.) form one row.
Additional columns — optional extra measures displayed next to the cohort size for context.
Step 3 — Set the Target measurement
The Target measurement section defines what's tracked across the columns:
Measurement field — the measure whose value you want to follow over time (e.g. active prospects, replies, revenue).
Measurement date — the date the activity happened. Alta computes how many periods passed between this date and the cohort date, and places the value in that column.
Step 4 — Adjust the General settings
Time interval — Daily, Weekly, Month (default), or Quarterly. Changing it automatically re-buckets the cohort date to match.
Value Type — Percentage (default; each cell is divided by the cohort size) or Absolute (raw values).
Value Format / Percent Format — number formats for absolute and percentage cells (e.g.
0.00%).Cumulative? — when on, each cell accumulates all previous periods instead of showing that period alone.
Step 5 — Optional: add a Summary row
In the Summary section, turn on Show summary? to add a row that aggregates every cohort per column. Choose the Summary type: Sum or Average.
Step 6 — Refine and save
Use Sort & Limit to control row order and count, Drilldown to define what opens when a cell is clicked, and Colors to override the color scale.
Click Save, name the visualization, and choose the dashboard where the widget should live — same flow as any Data Explorer chart.
How to read the chart
Each row is a cohort — all records that started in the same period.
Each column is the number of periods since the cohort started (period 0 = the starting period itself).
With Percentage on, a cell like 42% in column 2 means 42% of that cohort's size was still active (or did the measured action) two periods in.
Tips and common pitfalls
Percentages need a cohort size. If cells look wrong in Percentage mode, check that Cohort size is set — it's the denominator.
Expect a triangle shape. Recent cohorts haven't lived through as many periods yet, so their rows are shorter. Don't read missing recent cells as a drop.
Both date fields must be dates. The measurement date and cohort date are compared to compute the period offset; rows with an empty measurement date are skipped.
Gaps are filled with zeros. A period with no activity shows as 0, not blank, so trends stay readable.
Interval changes re-bucket automatically. Switching from Month to Weekly updates the cohort date grouping for you — no need to re-pick the field.
Related