There's a growing category of apps that offer a coach: something that checks in, asks how things are going, helps you set goals, and encourages you.
Some are good. What's worth having before you commit to one is a clear sense of which problems they solve, because these tools are genuinely good at two things, and the things that usually keep goals stuck aren't among them.
The two they're good at
Ambiguity, which is the most common reason goals stall and the cheapest to fix. If a goal is a vague noun — sort out the pension, get the business going — a language model will convert it into a list of physical actions in seconds, and sized reasonably. That's real work, and it's work people reliably fail to do for themselves, because deciding what starting means is the exact task that gets deferred.
Articulation. Writing out what's stuck, in enough detail for something to respond to, does some of the thinking on its own. And a reply — even a mediocre one — breaks the loop of turning something over privately. That effect is real and people who've felt it aren't imagining it.
Both of those are genuine, and together they cover a substantial fraction of stuck goals. Dismissing the category because it can't do everything is the wrong reaction.
The three they can't touch
Fear. A frightened person doesn't need a better plan; they have a perfectly good plan and can't approach it. Giving them a more detailed one adds another document to a goal that already has too many, and produces a small feeling of progress that displaces the actual move. Worse, the conversation itself becomes the safe activity — engagement with the goal that carries no risk.
Wrong goal. Whether dropping something would feel like relief or loss is not in your notes, and nothing in a conversation reliably surfaces it. Most of these tools are also built to help you achieve things, which means they'll help you plan a goal you should be deleting, enthusiastically, for months.
No capacity. When four unrelated goals stop in the same fortnight, the answer is to park most of them, and an app whose purpose is to keep you engaged with your goals is poorly positioned to tell you to stop having them.
That last point generalises. A product that measures its success by your continued engagement has a structural difficulty with advice like drop this, do less, or you don't need me this month.
None of that requires anyone to be acting in bad faith. It's simply that the incentives point away from a category of advice that people frequently need, and a tool whose revenue depends on your returning will find it hard to tell you convincingly that you shouldn't.
The relationship problem
The bigger issue with the coaching format is that it's a relationship, and relationships need maintenance.
There's a shape these apps tend towards where you end up performing — filing your day, reporting in, being told you're doing brilliantly. That's another thing to keep up with, and the people who need this most are already failing to keep up with things.
Then it becomes another source of guilt when you stop. Someone who's ignored their coach for three weeks has, in addition to a stalled goal, a stalled relationship with an app, and the second makes returning to the first harder.
There's also the encouragement problem. Consistent positive feedback from something that doesn't know whether you're doing well is worth very little, and after a few weeks most people stop reading it — which is fine, except that it was the product.
What a useful version looks like
Narrow and on demand.
One button, on one goal, when you press it. It reads that goal's notes and returns three moves of thirty minutes or less, plus a line quoted from your own record so you can verify it read them. Then it's finished — no check-in, no memory, no conversation.
The quote matters more than it sounds. It's a two-second test of whether the response is about your situation or about stalled goals in general, and generic advice about stalled goals is worth nothing.
And the plan-shaped answer is worth actively avoiding. A twelve-step programme with phases arrives looking authoritative and functions as one more thing to have and not do. Three small moves is a menu you can act on tonight.
The honest case for the full version
The counter-argument, because a narrow tool isn't obviously better for everyone.
Some people genuinely benefit from being checked on, and don't have anyone doing it. Working alone removes every external structure at once, and an app that asks how something is going reproduces a function that a job supplies for free — imperfectly, but not worthlessly.
Regular prompted reflection also works for some, and the fact that it's prompted by software rather than a person doesn't necessarily spoil it. Plenty of people have talked themselves into a useful realisation with something that was only nodding along.
And a well-designed coaching product with a subscription behind it can afford support, iteration and safety work that a thin free tool can't. There are worse things to pay for.
The narrower position is a preference rather than a proof: for goals that already matter to you, a tool you pick up and put down is less likely to become another obligation than one that expects you.
What it costs to run
One structural point that's easy to miss when comparing these.
Most of these apps hold an account with a model provider, route your text through their servers, and charge a subscription that covers the cost with margin. That means two parties see what you write, and it means there's a business that needs retention — which is where a lot of the check-ins, notifications and streaks come from.
The alternative is that you hold the account and the key stays in your browser. Requests go from your machine to the provider. Nobody in the middle sees the text, there's nothing to retain you for, and the cost of an occasional request is fractions of a penny rather than several pounds a month.
That's a worse experience to set up and a better arrangement for the notes you'd actually want to write.
Which is the trade in one sentence. Coaching apps optimise for the experience of using them. The narrow version optimises for what you're willing to write down honestly, and honest notes are where all the value in this practice comes from.
Related: AI can't want it for you · Bring your own key · What we send to the model, and what we don't
Share: