Where to find it | Sidebar → Client Health → Churn Log |
Who can see it | Everyone except standard-access users |
Can you change data here? | No — read-only |
How fresh is it? | Cached about 15 minutes |
What it's for
Signal has three views of churn, and mixing them up causes real confusion:
Page | Answers |
Who might leave — a model's prediction | |
Who account teams expect to leave — a human forecast | |
Churn Log (this page) | Who has left, and what it cost — the historical record |
This is the finance-style view. It's what you use to answer "what did we actually lose last quarter, and where?"
What you'll see
Summary cards:
Card | What it means |
Churned Clients | How many, with a note if any have a future churn date still pending |
Churned MRR | The revenue lost, at the final billed month |
Churned MGP | The gross profit lost |
Churn Rate | Churned revenue against the prior month's booked revenue |
Five tabs:
Monthly summary — the period-by-period picture
Division pivot — churn by division
Department pivot — churn by department
Churned clients — the list, sortable by churn date or by revenue at churn
MoM trend — month-over-month movement
Controls for granularity (month, quarter or year), period, division, department and client search.
How to use it
Report the quarter. Set granularity to Quarter, read the churned revenue and churn rate, then use the division pivot to say where it happened.
Find concentration. Compare the division and department pivots. Losses clustered in one department are a delivery conversation; spread evenly, they're a market conversation.
Rank the losses. On the Churned clients tab, sort by revenue at churn. The top few usually account for most of the damage, and they're the ones worth a post-mortem.
Check the trend. The month-over-month tab shows whether churn is accelerating. Read it against the ETCR page — churn on its own is only half the story if expansion is covering it.
Sanity-check a forecast. Compare what actually happened here against what Retention Forecast predicted for the same period. That's how you find out whether the forecast is trustworthy.
Good to know
Revenue is attributed to the client's last billed month, not the date someone recorded the churn. That keeps this page consistent with the finance view of when the revenue actually stopped.
"Pending" means a future churn date. The client is on the log because the churn is known and dated, but it hasn't happened yet. The summary card flags how many are in that state — don't count them as already lost.
Churn rate is measured against the prior month's booked revenue, so it's a share of what we had going in.
Clearing the period filter aggregates everything. Check what period you're actually looking at before quoting a figure.
This is the same page as the Churn Risk area in the navigation, but a genuinely different dataset. Churn Risk scores live clients; this records departed ones.
Common questions
Why doesn't this match Retention Forecast? Because they're different things. Retention Forecast is what account teams expect, typed into their tracker. This is what actually happened, from billing data. Comparing them is useful; expecting them to match isn't.
Why doesn't it match Churn Risk? Churn Risk only shows the active book — a client that churns disappears from it. This page is where they go.
What does "Pending" mean? A churn date in the future. Known and dated, but not yet effective.
Why is a client's revenue lower than I remember? It's their revenue in their final billed month, which is often lower than their peak.
Can I correct something here? Not on this page. Churn dates come from client records — fix them in Nova, or flag the number through Corrections.
Related
Churn Risk — who might leave next
Retention Forecast — what teams expect to lose
ETCR (Beta) — whether expansion covered these losses
Portfolio Review — the active book
Glossary — MRR, MGP, active book