Stop Writing Reports. Start Sending Recipe Boxes.

When I was at primary school, we were asked to write a synopsis of a book we had been reading.

I did exactly that.

The teacher gave me an excellent mark.

Then she realised I had copied the one off the back of the book word for word and adjusted my grade accordingly.

My argument was simple. The synopsis had clearly been written by professionals. There was no realistic scenario where I, as a 10-year-old, was going to produce something better.

You can imagine how well my defence went down. However, the teacher later told my parents she was genuinely torn. The initiative and thinking were hard to fault, and in a different lesson it would have been outstanding, but this task was about writing.

But it raises an interesting question.

What was actually being assessed? The quality of the output, or the thinking behind it?

That question feels rather relevant again now.

There was a time when a good piece of work looked like a finished document. Carefully written, well structured, and polished enough to feel complete. The kind of report that clearly took time and effort, and I still often produce these for the MATs and schools I work with.

Now, quietly and without much ceremony, that model has changed.

Because we all suspect what often happens next.

You send the report.
They open it.
And within seconds, it is being fed into an AI tool with the instruction: “Summarise this in bullet points.”

That moment matters more than it might seem.

Because it tells you something fundamental. The report you have just spent hours crafting is not the final product. It is raw material.

It is about to be reduced, reshaped, and repurposed into something else.

The document is no longer the end of the process. It is just one version of it, and often a temporary one.

Which raises a simple question. Why are we still spending so much time perfecting something that is about to be reshaped?

Perhaps the skill is no longer in producing the final version. Perhaps it is in sharing what sits underneath it.

Think about the difference between a takeaway and a recipe box.

A takeaway is built around immediacy. You decide what you want, when you want it, and it arrives quickly. Everything is designed for that moment. It is reactive, on-demand, and sometimes a bit rushed.

A recipe box, something like Gousto or HelloFresh, works very differently. You choose in advance. The company knows what you need days ahead. Ingredients are bought in bulk, prepared centrally, and delivered in a planned way. There is less urgency, less last-minute pressure, and far more predictability.

And when it arrives, you still have to do some work, but it is structured, guided, and manageable.

That difference matters. One model is built around constant, immediate output. The other is built around planning, structure, and shared effort.

That feels very close to the shift we are starting to see in how we work.

Instead of producing finished documents on demand, we can start sharing the ingredients earlier. Structured notes. Key data. Context. Clear intent. Here is what matters. Here are the boundaries. Here is what I need you to produce.

In effect, we are sending a recipe box rather than a finished dish.

And in many ways, we have already been moving in this direction.

Think about the rise of tools like Power BI across trusts. The whole point of those platforms is to move away from static reports and towards something more flexible. Instead of producing one version, you provide a dataset and allow different people to view it in different ways.

It is not quite a finished meal, but it is not completely open either. It is closer to a taco kit. You are given prepared components and a structure, and you assemble what you need within that framework.

But even that has limits. Dashboards are still designed in advance. Someone decides what questions can be asked, what filters are available, and what the views look like. It is flexible, but within boundaries.

AI takes this a step further. It is less like a buffet, where the choices are still predefined, and more like a grocery shop. You are given the ingredients and can produce entirely different outcomes depending on what you ask for.

That is incredibly powerful. But it also means the responsibility moves. The structure is no longer doing the thinking for you.

And this is where the shift becomes bigger than reports, dashboards, or even data.

It starts to change the nature of communication itself.

For years, we have focused on the presentation layer. Grammar. Punctuation. Structure. The way something reads on the page. These things matter because they make writing easier for people to consume.

But increasingly, what we produce is not being read in a fixed form. It is being interpreted, reshaped, and regenerated.

The same piece of content might become a summary, a set of bullet points, a presentation, or a set of actions, depending on what the reader asks for.

In that world, the original wording matters less than the clarity of the underlying meaning.

Punctuation and structure do not disappear, but their purpose changes. They are no longer just about presentation. They become signals. Ways of making intent clear, so that both people and tools can understand what matters.

We move from writing for reading, to writing for interpretation.

This becomes very real at certain points in the school year.

Think about what happens when data drops. GCSE results. SATs. End of term assessments.

There is often a familiar pattern. One person, or a very small number of people, suddenly becomes responsible for turning that data into everything everyone else needs. Reports for senior leaders, breakdowns for departments, summaries for governors.

Different formats, different views, different questions, but often the same bottleneck. Everything flows through the same place.

Even with dashboards in place, that reliance does not disappear completely. If something slightly different is needed, it still comes back to the same person to reshape it.

AI offers a different approach.

The data exists once, in a consistent and usable format that does not need time spent formatting it correctly. People have a small set of agreed starting points, and the ability to explore it themselves.

Instead of asking someone else to produce a new version, they generate what they need, when they need it.

The reliance on a single person reduces. The pressure spreads. The thinking improves.

And the role of the data lead shifts, from producing outputs to enabling others to use them well.

Which leads to a slightly different kind of question.

If communication is shifting from finished outputs to shared inputs, are we ready for that?

Because this is not just about efficiency or saving time. It changes how work flows through an organisation. It changes who produces, who interprets, and who is responsible for the outcome.

This is where an AI policy becomes important, but not in the way it is often framed.

A good policy is not just about what tools people can or cannot use. It is about how information is shared, how it is interpreted, and where the boundaries sit.

It should answer questions like: what should we provide as a finished output, and what should we share as structured input? What is safe to explore freely, and what needs tighter control? Where do we expect consistency, and where do we allow flexibility?

In other words, it is not just a technology policy. It is a communication policy.

Because if we get this right, we reduce bottlenecks, improve understanding, and make better use of the tools now available to us.

If we get it wrong, we risk increasing workload, losing clarity, and creating new dependencies in different places.

AI does not just change how we produce work.

It changes how work should move.

And perhaps the real opportunity for digital confidence is not simply adopting AI, but rethinking how we communicate so that it actually works in this new environment.

Now, if you will excuse me, I seem to have made myself hungry.