What decides whether it works?
- Whether it is actually used. Twenty minutes on most days beats an hour on Sunday, and a program a child resists produces nothing regardless of how good the content is. Short, finishable sessions are not a marketing choice; they are the whole mechanism.
- Whether it is pitched correctly. Practice at the wrong difficulty is either boring or demoralising, and both end in avoidance. Adaptivity per skill — not per subject — is what keeps a child at the edge of what they can do.
When is a human tutor the better choice?
- A diagnosed or suspected learning difficulty — dyslexia, dyscalculia, an attention difficulty. These need a person.
- More than about a year behind, with the gap not closing over a term.
- A child who will not work without an adult in the room. This is a real and common constraint, and no software solves it.
- An exam or transfer with a fixed date and a specific syllabus.
When is online practice the better choice?
- You are not sure whether there is a problem. Finding out is cheap; tutoring on a hunch is not.
- The gap is specific and known — times tables, inference, fractions.
- You want steady maintenance rather than intervention.
- Cost per child matters, especially with more than one child.
These are not exclusive. The common sensible pattern is daily online practice for maintenance and detection, with a tutor brought in for a defined stretch when the practice surfaces something that needs a person.