When half a cohort cannot attend, protect the women the programme was designed to serve, then rebuild the timetable around the constraint the data has exposed. A schedule is useful only when it gives participants a realistic chance to complete the work.
On Tuesday afternoon, I imagined Abena, a composite of the women such a programme might recruit, standing behind a pharmacy counter in Accra with her phone tucked beside the receipt printer. Her live session had already started. A customer was waiting, her supervisor was watching, and missing another class could cost her the place she hoped would help her move into technology.
At the same moment, the founder was looking at the attendance sheet. Half the cohort was absent.
The timetable said the programme was running. The data said access was breaking.
The timetable had become a product decision
The easy response would be to protect the calendar. Tutors had committed their time. Sessions had been planned. Moving one class could disrupt the next, frustrate the women who had attended and create more work for a small team already carrying too much.
There was also a harder possibility: perhaps the women who missed the session were less committed than the women who joined.
That explanation protects the programme from scrutiny. It turns an operational problem into a participant problem.
But attendance data cannot tell you why someone was absent. It records an outcome, not the constraint behind it. Work shifts, childcare, connectivity, device access and household responsibilities remain invisible unless the team goes looking for them.
The founder’s decision was therefore larger than whether to move Tuesday’s class. She had to decide which evidence defined success: a timetable delivered as planned, or women able to keep participating.
This distinction matters beyond education. A product can work exactly as specified while failing the people it was built for. I explored a similar measurement problem in What Happens When Signup Measures Access, Not Adoption?. Registration proves that someone reached the door. It says little about whether they could continue through it.
Absence is a signal, not an explanation
If one participant misses a session, the cause may be personal. If half the cohort misses it, I would first inspect the system.
That does not mean changing the entire programme after one difficult Tuesday. It means resisting the urge to explain the number before investigating it.
I would contact a small cross-section of attendees and absentees with three direct questions: What prevented you from joining? Which part of the current schedule can you reliably attend? What would help you catch up without lowering the standard?
The wording matters. “Why did you miss class?” can sound like an accusation. “What prevented you from joining?” treats the absence as information.
Then I would separate recoverable friction from structural conflict. A recording might help someone who lost connectivity for one evening. It will not solve a recurring clash between live sessions and paid work. A reminder might help someone who forgot. It will not create quiet space in a crowded home.
By Tuesday evening, Abena’s place would still be uncertain. If the programme required live attendance at a time she was routinely behind the counter, determination could not rescue her. She would have to choose between income and participation, and the programme would lose precisely the kind of woman it intended to reach.
Protect the outcome, then change the delivery
The standard should remain firm: participants still need to practise, submit work and demonstrate that they can apply what they learn. The delivery can be more flexible where flexibility does not weaken that outcome.
That could mean testing a second live window, publishing a catch-up path, repeating the most important working session or grouping participants by the times they can consistently attend. The right choice depends on what the conversations reveal. A founder should avoid designing a complicated solution before identifying the actual constraint.
The first change should also be small enough to reverse. Run it for the next session. Compare attendance, completed work and questions submitted afterward. Check whether the people who previously missed class can now participate, and whether the change has created a new problem for those who were already attending.
This is product work. A timetable is an interface between a programme and a participant’s life. When that interface repeatedly fails, discipline alone will not repair it.
The same principle applies when a signup flow technically accepts an account but blocks the person behind it. In Ama’s Failed Signup. A Customer Pilot Was at Risk., the meaningful question was whether the customer could reach the promised outcome before confidence disappeared.
Wednesday should produce better evidence
The founder does not need to declare the original schedule a failure. She needs to treat Tuesday as a warning that deserves a test.
On Wednesday, the useful artefact would be a simple table: the constraint each woman faced, how often it recurs, what recovery option she needs and which change the team will test next. No broad theory about commitment. No redesign built from guesses.
In Abena’s version of Wednesday, she finishes her shift, opens the catch-up task and sees a place in the next working session that she can attend. She still has to do the work. Her place in the cohort now depends on completing it, rather than being free at an hour she could never control.
That is the line worth protecting. Keep the learning requirement. Change the timetable when the timetable becomes the barrier.
Comments
No comments yet.