The situation
A headteacher is interested. Twenty students, two classes, a term. The temptation on both sides is to make it bigger and call it a launch.
The problem
A big trial of an early app gives a big pile of mixed results and a lot of upset parents. A small trial gives a clear picture. We would rather have the picture.
What we suggest
Decide what to try: the classes, books, language and the actual phones families use. Check that everyone is ready, including explaining it to parents and reviewing the rules on children's information. Let students ask real questions. Have teachers check the replies: are they right, are they clear, and was the right chapter found? Then review the experience together and decide the next step.
Watch for the things that matter: whether the right book was found, whether the explanation has the required facts, whether students understand the reply language, whether loading times are acceptable on the phones in the room, and where a teacher is still needed.
What we learned
The most useful trial feedback is not “students liked it”. It is “on these three phones it took over a minute to open a chapter” and “the Marathi replies in Geography were unclear”. That is what changes the product.
What it changed
The schools page lists these steps and these checks. It also says that none of them mean better results in school have been proved. We have not run a study that shows that, and we will not imply it.
Takeaway
Start small enough that one teacher can read every reply. If the trial is bigger than that, it is a launch wearing a trial's name.
The numbers here come from our release reports and one reported phone test. They do not show any change in marks or exam results.

