People did not abandon our five-day challenges because they were lazy. We built a format that competed badly with the rest of their working day.
Helen and I learned that after years of Collab365 conferences, workshops and online programmes.
The original article led with two internal figures: more than 50,000 people reached and roughly 7 per cent reaching day five of one Vacation Booking Challenge. Those are figures from our own operating history, not independently audited research. The useful point does not depend on treating them as a universal benchmark.
Many people started. Far fewer stayed to the end.
We blamed the content first
We tried the familiar fixes: shorter videos, more reminders, live sessions, community, office hours and extra encouragement.
Some changes helped. None removed the basic collision.
The participant still had email, Teams messages, children, clients, production incidents and a manager asking for five minutes. The programme required a clean stretch of attention that often did not exist.
There was a second problem in Microsoft 365. The product kept moving while the content took time to plan, record and publish. A polished walkthrough could begin ageing before a learner found the hour needed to watch it.
Completion was the wrong measure on its own
A completion rate tells you that somebody reached the end. It does not tell you whether the material solved their problem.
Equally, a person who watches one useful section and fixes their flow may have received more value than somebody who passively finishes every video.
That changed the questions we ask:
- What job is the person trying to complete today?
- What do they need to decide, build or check?
- How much context is essential before they can act safely?
- What would prove that the answer worked?
- When does the guidance need reviewing because the product changed?
Why we moved toward Spaces
We stopped treating a long programme as the centre of every answer.
Spaces is our attempt to organise current, problem-focused knowledge around the work somebody is actually doing. AI helps with research and maintenance, while human review protects the evidence boundary.
That does not mean every answer should take 15 minutes. The original article replaced one false rule, "long is good", with another, "short is good".
Length is not the standard. Fit is.
A security change may need careful prerequisites and testing. A keyboard shortcut may need three paragraphs. A strategic decision may need a proper comparison and time to think.
The hard-earned lesson is not that people have lost the ability to concentrate. It is that content earns attention by respecting the moment in which it will be used.
We still care whether people finish. We care more whether they can do something useful, safely, when the working day is already trying to drag them elsewhere.
Sources
- This article is a first-person account. Personal experience is not presented as independently audited external evidence.