Three parents ask what sounds like the same question in the same week. The first has an eight-year-old who reads chapter books happily but stalls on subtraction with regrouping: she gets the first two problems right, then the page goes blank and she says she is tired. The second has a nine-year-old who finishes the maths set in ten minutes and spends the other forty drawing on the inside cover of his exercise book. The third has a six-year-old who is not behind in anything, whose teacher has no concerns at all, and who does nothing between the end of school and bedtime except negotiate for more tablet time.
All three will search for something like "best learning app for kids" tonight. Only one of them has a problem an app can solve.
A tutor solves a diagnosis problem
When a child stalls on regrouping, the regrouping is usually not the problem. Underneath it there is normally something older and smaller: place value that never quite landed, or number bonds within ten that still have to be counted on fingers. If a child has to rebuild 13 minus 8 from scratch every single time, there is nothing left over for the borrowing step. She is not tired at problem three. She has run out of room to think.
Finding that is what a person in the room is for. A tutor can watch the child's hand, stop her mid-sum and ask "how did you get that one?", and change the plan in the next sentence based on the answer. A recorded lesson cannot, because it does not know which of six possible wrong moves this particular child is making.
Signals that this is your situation: the child was fine in this subject last year and is not now; the mistakes are consistent rather than scattered, the same wrong move again and again; she is cheerful about everything except one subject, which she avoids; or the class teacher has raised something.
If you do hire someone, hire them for a defined stretch. Six to twelve weeks, once or twice a week, with a named target such as "two-digit subtraction with regrouping, on her own, without counters." Ask for a diagnostic first session and a written note of what the gap actually is. Ask for ten minutes of homework, not forty. And talk to the class teacher first, because in most schools that conversation costs nothing and it sometimes ends the search on the spot.
Tutoring is the expensive option almost everywhere. As a rough guide, an hour with a qualified tutor costs about what one or two months of subscription software costs. That is not an argument against it; it is an argument for spending it on the thing only a person can do, and stopping when that thing is done. Small-group tutoring, two or three children sitting at a similar point, brings the price down and works perfectly well for a specific shared gap.
A course solves a "nothing new here" problem
The nine-year-old drawing on his exercise book is not stuck. He is under-fed. More practice of the thing he finished in ten minutes is the one response guaranteed to make it worse, because it confirms his theory that school work is a chore with no destination attached.
What he needs is a body of material with a beginning, a middle and an end: a course, a club, a teacher with a syllabus. There are two directions available. You can go faster, which means next year's curriculum this year. That sometimes works, and sometimes produces a child who is bored in exactly the same way twelve months later, now sitting with a teacher who knows he has seen it all before. Or you can go sideways, into something the school day does not cover at all: programming, chess, electronics, astronomy, an instrument, a language nobody in the house speaks. Sideways is usually the better bet, because depth is unlimited and speed is not.
The signals are distinctive: he finishes early and asks questions past the edge of the lesson, explains the topic to a younger sibling for fun, and calls it boring while resisting more of the same. He is refusing the format, not the subject.
Good courses have a visible sequence and something at the end. A program that runs. A game played at a rated club. A piece performed. A model that works. If in week six he cannot do something he could not do in week one, it was not a course, it was a subscription with a nice landing page.
A practice app solves a frequency problem
The six-year-old has nothing broken and nothing missing. What she has is no habit. Nothing happens unless an adult starts it, and the adult is tired at five o'clock.
This is the case where software is genuinely the right tool, and it is worth being precise about why. Practice software is good at a narrow set of things: turning up daily without a booking, starting in under a minute, keeping a session short, spacing out repetition of the material that only gets faster through repetition (letter sounds, number bonds, spellings, vocabulary in a second language), and leaving a small record a parent can glance at. For a five- to seven-year-old, ten to fifteen minutes is a session. Twenty-five minutes is a negotiation.
It is equally worth being precise about the limits. Practice software cannot work out why a child is wrong. It is a poor first teacher for a concept a child is genuinely stuck on. And it gives a bored child nowhere to go, because breadth is not depth. Sitting near a six-year-old while she works, not marking, just present, changes more than the choice of product does.
A quick way to tell which one you have
| What you notice | Most likely need | What that looks like |
|---|---|---|
| Was fine last year, is not now; the same error repeats | Tutor | Weekly, six to twelve weeks, one named target |
| Cheerful about everything except one subject, which she avoids | Tutor | Usually a gap plus lost confidence; both need a person |
| Finishes early, asks questions past the end of the lesson | Course | A sideways subject with a syllabus and an end product |
| Nothing is wrong; nothing happens unless you nag | Practice app | Ten to fifteen minutes, same daily anchor, you nearby |
| Missed six weeks through illness, moving country or changing school | Tutor first, then an app | A person to find what was missed, software to consolidate it |
| Learning the school's language as a second language | App for volume; tutor if speaking is the block | Daily exposure matters more than long lessons |
The mismatches that waste the most money
An app for a stuck child. More repetitions stacked on a broken step means practising the wrong method faster. The child collects a stream of red marks and concludes she is bad at maths, which is a far harder problem to undo than regrouping.
A tutor for a bored child. You are paying a premium hourly rate for supervised homework. He does not need supervision. He needs material.
A course for a child with no routine. Week one is enthusiastic, week three is a login nobody opens, and the unfinished twelve-week course sits there as evidence of something. Build the habit first with the thing that takes ten minutes, then buy the course.
All three at once. Twenty-five minutes of tutoring, twenty of a course and fifteen of an app is an hour of extra work after a full school day, for a child of seven. Something will be dropped, and it will be dropped in an atmosphere of failure. Choose one for this term.
What six weeks should show you
Give whichever you chose about six weeks, then check one specific thing rather than a general feeling.
- Tutor: the child gets further, unprompted, on the exact task she could not do at the start. Ask the tutor to state the before and the after in one sentence. If sessions have quietly become "we went through her homework," you are buying supervision.
- Course: something exists that did not exist in week one, and he can explain how it works without help.
- App: the session happens without an argument, and something surfaces away from the screen. She reads the sign in the car park. She works out the change before you do.
If the answer is no, change category before you change brand. The most common wrong move is swapping one app for another app when the child needed a person, or one tutor for another when the child needed new material.
It is also normal for the answer to change. The eight-year-old may need a tutor for one term and then nothing but ten minutes a day for the following year. And two siblings in one house rarely need the same thing in the same month, which is worth saying out loud before you buy one plan for both of them.