Client case study · 6 months Illustrative

23%→61%

Learners still active at day 60, over 6 months.

Elapsed: 6 months
How it was measured: Cohort retention measured at day 60 across consecutive intakes, same acquisition source.
Running alongside: No change to content or pricing in the period.
Published with permission: Anonymised at client request

The constraint

A plan that breaks the first day you miss

A fixed study plan assumes a life without night duties, family events or illness. Miss two days and the backlog makes the plan feel unrecoverable, so people stop opening it.

The drop-off was concentrated in weeks three and four, right after the first missed block.

Fixed schedules with no recovery path
Revision by calendar date rather than by what was actually forgotten
No signal to the learner that they were still on track

What we found

The audit, including the unflattering parts

We looked at where cohorts actually stopped.

77% inactive by day 60
Median abandonment three days after a missed block
Revision items ignored at a much higher rate than new items

The plan

The 90-day roadmap as it was actually written

Reschedule around real life; revise by decay, not by date.

Missed days redistribute across remaining capacity automatically
Revision surfaces by forgetting curve rather than fixed interval
A hard daily cap so a bad week never produces an impossible day
Weak topics detected from performance, not self-report

Month by month

What shipped, and what it moved

The line extends as you scroll. Every step below is a real month, including the one where nothing happened.

01Month 1

Baseline

77% of learners inactive by day 60, most quitting within three days of a missed block.

Baseline: 23%

02Month 2

Adaptive rescheduling

A missed day redistributes instead of accumulating.

+17%

03Month 3

Revision by decay

Items surface when they are likely forgotten, not on a fixed date.

+26%

04Month 4

Daily cap added

After the failure below, no day can exceed a manageable load.

+32%

05Month 5

Weak-topic detection

Performance data drives focus rather than self-assessment.

+20%

06Month 6

Where it landed

61% still active at day 60, from 23%.

+165% total

What went wrong

The part most case studies leave out

The first revision algorithm pushed everything a learner had ever marked weak, which produced 200-item days and made abandonment worse before it got better.

Our first revision algorithm made the problem worse.

It surfaced everything ever marked weak, which produced days with more than two hundred items. Learners opened it once, saw an impossible list, and left. Retention dipped before it recovered.

The fix was a hard daily cap with prioritisation — show the twenty items that matter most today, not everything outstanding. Any adaptive system needs a ceiling, or it punishes exactly the people it is meant to rescue.

The numbers

Where it landed, and how we know

Learners still active at day 60: 23% → 61% over 6 months. Cohort retention measured at day 60 across consecutive intakes, same acquisition source.

Cohort retention at day 60 across consecutive intakes from the same acquisition source. Content and pricing unchanged.

Active at day 60: 23% to 61%
Median session length: up 2.4x
Revision completion: 31% to 78%

Results reflect this client's market, licence status and starting position. We publish the method alongside the number so you can judge whether it transfers to yours.

In hindsight

What we'd do differently next time

Cap the daily load from day one
Ship rescheduling before revision; recovery matters more than optimisation
Show the learner they are on track — the reassurance retained as well as the algorithm did

Before you ask us to do this for you

Does this actually apply to your situation?

This worked because of specific conditions. If yours don't match, say so on the first call and we'll tell you what would change.

Market: Any long-horizon learning product where consistency decides the outcome.
Starting point: Existing content and enough learners to measure a cohort.
Timeline: Five to eight months, because retention takes cohorts to measure.

Start with the audit. It's free and it's specific.

Send us a URL. You'll get back the technical issues, your AI-search readiness score, what competitors are doing that you aren't, and what we'd fix in the first ninety days.

Get a free audit → WhatsApp us instead