Product Features Pricing FAQ
← Portfolio
Predictive churn intelligence

Know who's leaving⁠— before they do.

PULSE watches every signal your customers give off, predicts churn weeks ahead, and tells your team exactly what to do about it.

Start Free Watch Demo

No credit card required · 14-day trial · Cancel anytime

Trusted by revenue teams at

01 — Predict

See churn
weeks early.

PULSE scores every account daily against 240+ behavioural signals — login decay, feature abandonment, support sentiment, invoice friction. You get a ranked list of who's slipping, long before they open a cancellation page.

  • Daily risk scoring across your entire book
  • Signals from product, billing, and support in one model
  • Alerts routed to the owner who can act
Accounts at risk Live
NW Northwind Labs
$18.4k ARR · 62% usage drop 91
HX Helix Digital
$9.1k ARR · seats unused 67
AT Atlas Grid
$24.7k ARR · 2 failed invoices 58
VR Verdant Co.
$12.2k ARR · healthy 14
CB Cobalt Systems
$84k ARR · recovered 08

02 — Explain

Know exactly
why.

A score without a reason is just anxiety. Every PULSE prediction unpacks itself — the precise signals driving the risk, weighted and ranked, in language your CS team can read out loud on a call.

  • Per-account attribution, not black-box scores
  • Plain-English summaries generated for every alert
  • Compare against accounts that looked identical
Why Northwind Labs is at risk 91 / 100
Weekly active users−62%
Core feature abandoned34 days
Support sentimentNegative
Champion left companyConfirmed
Invoice friction1 retry

03 — Prevent

Fix it before
it breaks.

PULSE turns each prediction into a playbook — the specific action, for the specific account, with the expected lift attached. Approve it and the work runs itself across the tools you already use.

  • Playbooks ranked by predicted recovery rate
  • One-click execution into your CRM and inbox
  • Every outcome fed back into the model
Recommended playbook Running
Flag account to owner done
Send re-onboarding sequence +41%
Book executive check-in +23%
Offer annual migration +18%
Escalate to retention pod +9%

0%

Prediction accuracy

0

Teams

$0M

Revenue saved

0M

Predictions processed

The product

Your whole retention picture, alive.

Not a screenshot. This is the interface, running — updating, predicting, and writing its own recommendations while you read this.

app.pulse.example/retention

Retention overview

Last 30 days · auto-refreshing

Live

At risk

38

↓ 6 today

Recovered

129

↑ 12 this week

ARR protected

$1.84M

↑ 4.2%

Predicted vs. actual churn Predicted Actual
A laptop on a bright desk running the PULSE analytics dashboard, surrounded by floating interface panels. Running on its own
The flow

It works while
you don't.

PULSE connects once, then disappears into the background. No dashboards to babysit, no weekly export ritual. The hard part — noticing — is handled.

  • 01

    Connect your stack

    Product events, billing, CRM, support. Ten minutes, no engineer required.
  • 02

    PULSE learns your shape

    The model calibrates to how your customers actually behave — not an industry average.
  • 03

    Act on what surfaces

    Your team wakes up to a ranked list and a recommended move for each account.
Pricing

Priced on what you keep.

Every plan includes the full prediction engine. You scale seats and volume, not features.

Monthly Annual Save 20%

Starter

For small teams protecting their first thousand accounts.

$49 / month

Billed monthly

Start Free
  • Up to 1,000 tracked accounts
  • Daily risk scoring
  • 3 integrations
  • Email alerts
Most popular

Growth

For revenue teams running retention as a real motion.

$149 / month

Billed monthly

Start Free
  • Up to 25,000 tracked accounts
  • Full attribution & explanations
  • Automated playbooks
  • Unlimited integrations
  • Priority support

Scale

For platforms where a single logo moves the forecast.

$429 / month

Billed monthly

Talk to us
  • Unlimited tracked accounts
  • Custom model tuning
  • SSO, audit logs, data residency
  • Dedicated success engineer
  • 99.9% uptime SLA
FAQ

Questions,
answered.

Everything teams ask before switching their retention motion over to PULSE.

Most accounts surface 3 to 6 weeks before they would have cancelled. The window depends on your contract length and how much product signal you send us — high-frequency products give the model more to work with, so warnings arrive earlier.
About ten minutes for the connections, then roughly a week of quiet calibration while the model learns your baseline. You'll see your first scored list the same day, and it sharpens from there.
No. PULSE ships with pre-built connectors for the tools most teams already run, and every prediction arrives with a plain-English explanation. If you do have a data team, the raw scores are available through the API.
It stays yours. Data is encrypted in transit and at rest, never used to train models for other customers, and deleted on request within 30 days. Scale plans can pin storage to a specific region.
That's the intent. PULSE isn't trying to replace where your team works — playbooks execute into your CRM, inbox, and messaging tools so the recommendation shows up where someone will actually see it.
It will sometimes — 94% accuracy means roughly one in sixteen calls misses. Every outcome you record feeds back into the model, and the attribution view shows you exactly which signal misled it, so wrong answers are at least legible ones.
Start today

Stop finding out
on the way out.

Fourteen days, your real data, no card. If PULSE doesn't surface an account you'd have lost, walk away.

Start Free Watch Demo

Set up in 10 minutes · Cancel anytime