Read the numbers on each step
What entered, here now, left and the completion rate actually count.
Each step on the canvas carries its own figures, and there’s a flow-level summary above it. They’re worth reading precisely — a couple of them don’t mean what the word suggests.
On each step
| Figure | What it counts |
|---|---|
| Entered | Everyone who ever reached this step |
| Here now | People standing on it at this moment |
| Left | People whose run was cancelled while on it |
| Opened / clicked | For email steps, on the mail that step sent |
“Left” is not everyone who moved on
It counts people who dropped out — a cancelled run — not people who continued to the next step, and not people who finished the flow. Reaching the end is a success, so counting it as a loss would put the alarming number on the step that worked.
The summary above the canvas
- Entered — runs that started.
- In the flow right now — live, this instant.
- Finished — runs that ended, by reaching the end or hitting a goal.
- Dropped — runs cancelled.
- Completion — finished as a share of runs that ended.
The worst drop-off step is named for you, which is usually the fastest way to find the email that isn’t working.
Why the date range doesn't apply to everything
You can narrow the page to a date range. It applies to some figures and deliberately not to others, because there are three different kinds of number here:
- Events scope cleanly. Someone reaching a step, a run starting or ending, an email going out — all happened at a moment, so a window includes them or doesn’t.
- Current state ignores the range. “Here now” describes this instant. Filtering it by a window that closed last month would render zero and read as “nobody is in this flow” — a lie about live data. The labels say which is which.
- A/B standings ignore it too. They’re the evidence a winner was decided on, and showing a slice of that evidence next to a winner computed from all of it would be a contradiction.
Entered and finished are different groups
Inside a window, someone can finish who entered before it started. So the completion rate can exceed 100% on a narrow range. That isn’t a bug — it’s two different cohorts, and the page says so.
The range defaults to the flow’s whole lifetime rather than the usual 30 days: a flow runs for months, and silently switching every figure to a month would change numbers you’d already read without anyone asking.
Email steps count by when the mail was sent
An email step’s figures scope by when it sent, not when someone opened. Mail sent inside the window is what that step is responsible for; counting an open of a months-old message would credit today’s window for old work.
A flow with no email steps shows dashes rather than 0% — a rate over nothing isn’t zero.
Where else to look
The Automations tab in Analytics rolls every flow up in one table, so you can compare them before opening any one. Welcome emails appear there too.
Automation mail is kept out of the other analytics views on purpose: a flow has one subject line and months of sends, so folding it into per-campaign averages would let one drip dominate every number.
Common issues
“Entered” is higher than the number of people I have
Repeat entry is on and some people have been through more than once. These count runs, not people. See Letting people come round again.
A step shows people entered but nothing after it
They’re probably waiting, or a branch below it isn’t wired. Check “here now” on that step.
Completion is over 100%
A narrow date range where more runs ended than started. Widen it, or use “Show all time”.
An email step shows no opens or clicks
Tracking arrives over minutes. If it stays empty on a step that has definitely sent, check the campaign report behind that step.
FAQs
Do these count people or runs?+
Runs. With repeat entry off they’re the same thing; with it on, one person can account for several.
Why is “left” zero on my last step?+
Because reaching the last step and finishing is success, not drop-off. Only cancelled runs count as leaving.
Can I export this?+
Per-step figures live on the canvas. For exportable data, each email step has a full campaign report behind it, and Analytics exports CSV.