Claim and evidence
3.36% drop in trial cancellationsMonarch Money · customer.io · undated
Vendor claim
A tool vendor reports numbers about its own customers or product. Useful, but not independent.
3.36% drop in trial cancellations.
Working · 2026-11 Customer.io case study (vendor) and Heap's own blog (2021).
What the numbers show
drop in trial cancellations after switching to behaviour-triggered emails (Monarch Money, vendor case study)
customer.ioconversion to sharing a report among users who opened the email (Heap; openers only, not a controlled comparison)
heap.ioWhy it works
Messages arrive when relevant and are sent only to the people who need them.
How to run it
- Define behavioural cohorts (installed but no event, etc.)
- Write one short series per cohort
- Sync segments to the email tool
- Use a holdout group to measure
Before / after
Generalised wireframe of the pattern, not a copy of any company's screen.
Ways to run it
Design options to test, not promised results.
- Option ACohort series
Write a short series for each behavioural cohort, such as installed but no event, as Heap did.
- Option BReplace the trial drip
Swap generic time-based trial emails for triggered ones and validate with a holdout, as Monarch Money did.
- Option CSegment sync
Sync behavioural segments from product analytics to the email tool so messages follow actual usage.
Timeline
- 12021 Heap — Cohorts by actions taken; segments synced to Marketo.
- 22026-11 status checked: Working
Dated examples
- 2021
Heap
Cohorts by actions taken; segments synced to Marketo.
Those who opened the email converted 10% more.
“Overall, from signup to share a report, those who opened the email converted 10% more!”
✓ quote checked word-for-word on the pageheap.io · 2021-05-07
- —
Monarch Money
Replaced four generic time-based trial emails with behaviour-triggered messages, with holdout tests.
3.36% drop in trial cancellations.
“3.36% drop in trial cancellations”
✓ quote checked word-for-word on the pagecustomer.io · undated
Does it still work?
Customer.io case study (vendor) and Heap's own blog (2021).
Risks
- Heap's lift is for openers, which is self-selected
- Vendor-published
Sources
- customer.io
“The team also used Customer.io’s built-in holdout testing to validate the impact of their new approach.”
✓ quote checked word-for-word on the page - heap.io
“With behavior-based emails, you segment your users into behavior-based cohorts”
✓ quote checked word-for-word on the page
Change log
- 2026-11-20 — harvest v0.1 (activation)