There is a number that gets reported at the end of a device rollout, and it is almost always the wrong one.
Ninety-two percent delivered. Four thousand units shipped. Ratio moved from one device between three students to one between two.
Every one of those is a logistics number. Not one of them tells you whether a single lesson changed.
the order things happen in
Here is the sequence I keep walking into.
Somebody senior decides the organisation needs to close the device gap. A budget appears. Procurement is asked how many, how fast, and how cheaply. A central team works out the shortfall per site, subtracts what is already there, and places the order.
The devices arrive. Everyone takes a photo. There is a press release.
And then, months later, someone asks the question that should have come first: what are the teachers actually doing with them?
the tell
When I want to know whether a rollout worked, I do not look at the delivery report. I walk into a classroom.
I want to know what the room looked like before the devices arrived, and what it looks like now. Is the teacher still at the front? Is there still a queue at the photocopier? Are children still being handed twenty-five sheets of paper that will be recycled by Friday?
If the answer is yes, you did not buy a learning tool. You bought furniture. Expensive, fragile furniture with a four-year life and no plan for what happens at the end of it.
what goes wrong, specifically
It is rarely one big failure. It is four small ones, and they compound.
Nobody asked the school what it needed. The gap was calculated from a spreadsheet. So a primary that teaches beautifully on tablets receives laptops, because laptops were what the volume discount covered. There is usually a perfectly good rationale for the decision. There was just never a conversation. And a decision without a conversation is an imposition, however sound the reasoning.
Nobody costed the environment. The devices land in a building with no plan for where they charge, no cupboard that locks, wifi that was specified for half the traffic, and a licence that has not been bought. The kit is not wrong. It has simply been set up to fail, and then people are surprised when it does.
Nobody asked what the teaching was supposed to become. This is the big one. If you cannot describe what a good lesson looks like with the device in the room, you cannot brief a teacher on how to use it, you cannot tell whether it worked, and you certainly cannot justify buying more next year.
And then the people who were not consulted become the problem. They get called resistant. They are not resistant. They were handed an object and no reason, and they did the reasonable thing, which was to carry on teaching the way they already knew worked.
why it keeps happening anyway
I want to be fair about this, because the pedagogy-first argument is often made by people who have never had to spend a budget before a fiscal year closes.
Pedagogy-first is slower at the start. It means walking classrooms, talking to teachers, running a discovery phase, and holding your nerve while the money sits unspent. Meanwhile the prices are moving, the chips are constrained, the lead times are stretching, and someone above you is asking why nothing has been ordered.
So the device decision gets made first because it is the one that can be made first. Not out of stupidity. Out of pressure.
The cost lands later, and it lands somewhere else. On the teacher who never got the training. On the child whose lesson did not change. On next year's budget, which now has to justify itself against a programme nobody can prove worked.
what actually helps
Not a slower process. A differently ordered one.
Find what is already brilliant before you change anything. Somewhere in the organisation there is a department doing genuinely good work with technology. Start there. It gives you a model that already fits your culture, and it means the first thing people hear is not that they have been doing it wrong.
Ask schools for their requirements, not their solutions. "What do you need to be able to do?" gets you something you can design against. "What do you want to buy?" gets you a wish list and a fight.
Decide what a lesson should look like, then buy the thing that makes it possible. Written down. Specific enough that a teacher could recognise it, and that you could tell six months later whether it happened.
Phase it. Ten schools, not a hundred. Learn on the ten. The instinct to do everything at once is what turns a fixable mistake into a hundred copies of a fixable mistake.
Put safeguarding at the front. I would not have said this a few years ago. I do now. Every device you hand a child is a door, and safeguarding is the one lens that makes every other conversation, about filtering, about training, about which platform, about what data goes where, line up behind a single answer everyone already agrees with.
the thing I would actually say in the room
You will get one honest run at this. The money is provisional until the plan is good, and the goodwill of the people who have to teach with it is finite.
So spend the first part of it finding out what is already working, and the second part making the case for what teaching becomes. The devices are the last decision, not the first.
Delivered is not adopted. And adopted is not yet a child learning something they could not learn before.
If you are partway through a rollout and quietly worried it is not landing, that is usually the right instinct and it is not too late to change the order. I do readiness audits for exactly this.
