The Pulse dashboard
How to read the main project dashboard, and exactly how each number is calculated.
The Pulse is the first screen you land on for a project. It answers one question: is this week better than the last one, and where is it going wrong?
Controls
Two controls sit at the top.
Range selector — Today · 7d · 30d · 90d · 12mo. The default is 30d.
Local / UTC toggle — switches how times are rendered. It is display-only and stored in your browser (localStorage), and it applies instantly across the Pulse, Sources, and Events pages. Query bucketing is always UTC, so when you’re in Local mode the hour and day labels can look shifted relative to where the buckets actually start. That’s expected — the toggle changes labels, not math.
The stat strip
The strip adapts to what you’ve configured in project settings. You always get:
- Visitors, with a percentage delta against the immediately preceding window of equal length (↑ or ↓ n%). If the previous period had zero visitors, the delta is omitted entirely rather than shown as an infinite jump.
If you’ve configured a conversion, you also get the converter count — labeled with your event label if you set one, otherwise “Signups” — its own delta, and a hero Conversion % tile.
If you’ve configured a first-value event, you get Reached first value and Typical time to value. Time to value is a median, formatted as Ns under 90 seconds, Nm under 90 minutes, and Nh above that. If you have not configured a first-value event, a Revenue tile takes that slot instead.
Charts
“Visitors over time” buckets by hour for today, by month for 12mo, and by day for everything else.
When you have a conversion and/or a first value configured, two smaller stage charts appear underneath: conversions in orange, first values in green. They use the same buckets, so you can line up a traffic spike with whether it actually converted.
Journey funnel
A single panel reading left to right:
Visitors → (conversion rate %) → converters → (activation rate %) → activated.
This is the whole product in one line. Traffic that doesn’t convert is a landing page problem; converters who don’t activate are an onboarding problem.
Sources panel
Your top 5 sources by visitors: Source / Visitors / Conv. % / FV %. The FV column only appears when a first-value event is configured. “View all sources →” takes you to the full Sources page, which shows the top 20 with landing pages and countries.
Live activity
A full-width dark panel streaming events as they arrive, with columns Event / Who / Path / Source. It keeps the last 8 rows. Event names ending in _created render green.
The live stream only carries events that arrive after you connect, so the panel is seeded with the last 8 events from today. That way it isn’t blank when you open the page at a quiet moment. For a complete, queryable record, use the events log.
Exact metric definitions
Do not guess at these. They’re precise, and knowing them saves you from misreading a chart.
| Metric | Definition |
|---|---|
visitors | Count of distinct visitor_id |
pageviews | Count of $pageview events |
converters | Distinct actors with at least one event matching your conversion definition |
activated | Actors whose first-value timestamp exists and is at or after their conversion timestamp |
time to value | Median of (first value time − conversion time) across activated actors |
conversion rate | converters ÷ visitors × 100, to one decimal place |
activation rate | activated ÷ converters × 100 |
Note the last row carefully. Activation rate is a share of converters, not of visitors. A 40% activation rate means 40% of the people who signed up got to first value — not 40% of everyone who showed up.
Actor vs visitor
This distinction explains most confusing numbers on the page.
An actor is the user_id when you’ve set one via identify(), and the visitor_id otherwise.
So “Visitors” counts distinct visitor ids, while converters and activated count distinct actors. If a person browses anonymously, signs up, and then logs in from a second device, they can be two visitors but one actor. That’s the correct behavior — you want the funnel to follow the person, not the browser.
Range semantics
| Range | Window | Buckets |
|---|---|---|
today | Since UTC midnight | Hour |
7d / 30d / 90d | Rolling N×24h from now | Day |
12mo | From the 1st of the month, 11 months back | Month |
Deltas always compare against the immediately preceding window of equal length. On 7d, that’s the 7 days before the last 7.
Note that 7d/30d/90d are rolling windows from the current moment, not calendar days. If you load the page at 3pm, the 7d window ends at 3pm today and starts at 3pm seven days ago.
Caching
Dashboard data is cached briefly — overview and timeseries around 30 seconds, sources/pages/countries around 60 seconds. The cache key includes your project configuration, so if you change your conversion or first-value event, the change takes effect immediately instead of waiting out the cache.
Where to go next
- Sources — which channels actually bring people who convert
- Events log — the raw stream, filterable and shareable
- Weekly Pulse — the same story delivered to your inbox
- Conversions and activation — how to define the two events this page is built around