You log breakfast, you log lunch, and the scale has barely moved for weeks. Then the app offers to put a person on the other end of all that logging, for a price.
Researchers have tested almost exactly that offer. In the biggest of those trials, by its own report, the extra wasn't a person: it was messages a computer program chose.
Think of a forecast that says probably dry while its range still covers a shower. That's the coach upgrade, measured. In five trials of the same kind of weight-loss app, with or without a coach on top, the coached groups came out 1.80 kg ahead on average at the first check after the programme. The far edges of that estimate run from a solid few kilos better to slightly worse.
The review that brought those trials together, led by Chew and published in 2023, couldn't tell that gap apart from chance. It's a lean, not a result.
The same review also reported a bigger number, just over two kilos, and that's the one that travels. It mixes in people who had no app at all. Against them, the coach-plus-app package came out 2.27 kg ahead. That's a real gain, but for the whole package, so it can't say how much was the app and how much the coach.
That usual-care comparison is also where the review's two automated programmes sat; neither was among the five. Only the comparison of a coach against the same kind of app matches the button on your screen.
Who was on the other end?
The five head-to-head trials have a blur of their own. By its own report, the biggest gave everyone a single session with a dietitian, then sent one group messages a computer program picked. The second-biggest gave both groups the same counsellor-led programme and changed only the app. Together they hold 566 of the 700 people in that comparison.
Even the research can lose track of the person. A newer, larger analysis counted computer-sent emails carrying a dietitian's name as contact with a specialist.
Before asking whether a coach is worth it, the evidence asks what the coach is. If you've wondered whether the coach in your app is really a person, the research ran into the same question.
- Biggest trial, 502 people. Everyone had one session with a dietitian. Then one group also got messages a computer program picked.
- Second-biggest, 64 people. Both groups had the same counsellor-led programme. Only the app changed.
- The other three trials, 134 people.
What the nudges moved
The extra support barely shifted the scale. What it did shift was mostly behaviour. In the biggest trial, the people getting the computer-picked messages hit their calorie goal on more days, and at a year they had lost no more weight.
Picture an inbox of nudges, most of them left closed: the typical person in that group opened fewer than half of what arrived. The trial's own authors wrote that without personal interaction, engagement fell away. They didn't test a person in its place.
Staying with it is what a coach is meant to sell, and the other study this answer rests on counted exactly that, in lifting rather than weight loss. In the lifting trial, run by Gavanda's team, more people coached through an app reached at least 26 of the plan's 30 sessions than people handed a PDF: 81% against 52%. The trial never checked whether that gap could be chance.
Among those who did, the app lifters gained about the same squat strength as the PDF lifters. Lifters with a coach standing beside them put about 7 kg more on their squat than the app lifters. That coach was catching each rep in the room, a different job from reading a food log.
Support showed up in what people did, more than in what the scale said. Whether a person keeps people going in a way that adds kilos wasn't tested in either study. It's a hint about staying power, not a promise. It is a hint the wider evidence repeats: almost every kind of paid help showed up first in who stayed to the end.
Would more coaching help?
Maybe the fix is more: more sessions, more contact, a coach you can see. Every time the research we examined stacked support on top of support, the gap came out close.
In the review, the number of coaching sessions showed no link with how much weight people lost, though several trials reported their sessions poorly. A review led by Hartmann-Boyce, of trials comparing more contact with less over the same period, found a 0.3 kg difference at a year.
Seeing the coach in person didn't open a gap either. In a two-year trial in adults with obesity and a heart-disease risk factor, coaching by phone, website and email ended 0.5 kg apart from the same kinds of remote support plus face-to-face sessions. The gap was too small to call, and that remote coaching wasn't an app.
Picture two steps on a staircase: the first, from nothing to an app with a coach, is the one the trials could measure. The second, from the app to the app plus a person, is the one they couldn't.
What these trials never tested
None of the trials we examined tested what your upgrade screen is selling: a paid coach tier inside a commercial app, set against the same app without it. The coaching here was research programmes, run mostly by dietitians and psychologists; one of the five used a commercial coaching app, with psychologists as the coaches.
The review used each trial's first check after the programme, between 2 and 24 months in, in adults with overweight or obesity around 39 on average, without long-term illness or mental health conditions. The lifting trial ran 10 weeks, in people who already lift.
That's why this answer carries a low-certainty rating: few head-to-head trials, a wide range, and trials where the person often wasn't what was being compared.
Before you tap upgrade
Put together, the evidence lands in one place: paying for a person on top of an app you already use is an upgrade these trials never showed on the scale.
If you're starting from nothing, the package of an app plus a coach did beat usual care, without showing which half did the work. Tracking on its own has its own evidence, from the 12 trials of digital food and exercise tracking on our tracking page.
Two things decide whether the tier is even the thing these trials hinted at, and only you can check them. One is whether the coach is a person who reads your log or messages a program sends. The other is whether you'd actually answer it.
If you know you drift off apps after a few weeks, that's the one case where these studies lean toward a person, gently and untested. If an AI coach is tempting you instead, not one of the studies here set one against a human.
If keeping you going is a person's real job, it's fair to ask what else can do it. In a trial led by Ladapo, everyone already had the person. Its 668 primary-care patients with obesity, in low-income neighbourhoods, all got a year of WW, help joining its groups and a one-on-one check-in every month.
The money was what changed, though its authors note the cash groups also got feedback on their incentives, and so perhaps more human contact. This time the result is counted in people, not kilos. With one of two cash designs added, the share who lost at least 5% of their weight by six months went from about one in five to about half.
Your upgrade price pays for the one ingredient the head-to-head trials couldn't size: the person. Researchers feel that cost too. The team behind the second-biggest trial called human counselling the main expense of app-based weight-loss programmes. It's also what stops those programmes from reaching more people. They wrote, too, that it goes with better results, though in their own trial both groups had a counsellor, each its own, so it didn't test that.
What the trials saw more clearly was who kept going, and that's the part only your own habits can answer.