There is no way to write this without sounding as if I am bragging: every morning, I wake up and do a 12-minute mini-workout, shower, then learn some Japanese over breakfast.

I know, I know. But I am also one of an increasing number of people who have turned to apps to make themselves fitter, healthier and more productive – or, at least, to accrue a few more good habits. Take Ed, a writer from London, who has lost 70kg (11st) using calorie-counting app Lose It!; Gareth, a developer, who credits the Drink Free Days app with helping him to get a handle on his drinking; and Sarah, from Newcastle, who says her period-tracking app has helped her to predict PMT.

Health and fitness-focused apps are wildly popular, and studies have shown that they can have a real impact. Research by Flurry mobile analytics found that usage of such apps grew by 330% between 2014 and 2017, and that more than 75% of active users accessed their app at least twice a week. More than 25% opened them at least 10 times, reflecting high engagement and, the analysis suggested, high retention rates. The NHS has even approved some apps to help with problems such as anxiety, self-harm and depression.

But for every success story, there seem to be as many failures. I have tried calorie counting and found that, after a couple of months, I stopped using the app. Even the Flurry research, which found people using health and fitness apps to be “the most loyal users in the app industry”, noted “fluctuating engagement levels”.

Some apps are so irritating that they have the opposite effect to that intended. Agnes, who used an app to remind her to drink water, says: “It pissed me off so much that I drank less water than before.” Worse still, they can be too effective. Madeleine says calorie-counting apps played a part in her developing an eating disorder: “Central to recovery, after losing dangerous amounts of weight, was deleting [them].”

The very habits that apps seek to cultivate – rigorous self-assessment, competition and a single focus on bringing down one number – have the potential to become self-destructive. Many aim to build a habit; in a minor way, rewiring your brain. But what makes some apps useful and others useless? And how much of that has to do with the app itself, and how much is to do with how it is used?

My morning workout is courtesy of Streaks Workout, which shows me how many days each month I have managed to fit in some training. It is a spinoff of Streaks, another app by the same developer, Crunchy Bagel, which aims to help users build good habits or ditch bad ones.

Quentin Zervaas, the company’s founder, says the secret is breaking the long-term rewards of successfully keeping up a habit into more immediate reinforcement: “If you’re trying to do anything, such as losing weight, it takes so long before you see results. You don’t get a reward straight away, but you don’t get punished either. With an app such as Streaks, you get the feedback instantly – it’s about being conscious, being aware of doing something.”

He continues: “Someone at Apple told me that something you need to look for is the moments of delight – if you can give the user moments of delight, that’s what keeps them involved.”

Tap an icon and feel a sense of achievement and maybe you will be more likely to do it the following day; keep doing it and, eventually, you will not need the app at all. That is the goal of all these apps – so why do some fail? A big problem, it turns out, is that many of the habits they are attempting to build are not the habits users think they are building.

Getting into the habit of working out every morning is, by and large, a good thing; getting into the habit of logging it isn’t, necessarily. Declan’s experience of MyFitnessPal chimes with mine. He says: “After a while, it stopped giving me anything of real use and began to feel like something tacked on at the end of the day. My goals were just kind of there – I could either complete them or not.”

If there is one distinguishing feature of people who have had success with these apps, it is that they consciously and deliberately link real-world improvement with the habits the app is building. When I used MyFitnessPal, I very quickly built the habit of logging my food into the app – but I didn’t change my diet.

Contrast that with the experience of Ed, who said that LoseIt!’s non-intrusive reminders helped him to lose that 70kg. “If people are actually honest and accurate, logging will make them lose weight. Logging automatically makes you eat less, as you’re far more aware.”

It is possible that MyFitnessPal – which invites you to dig into nutrition reports, pay for a premium version and share your diet with friends – may be asking too much. After finding other apps too clunky and not compelling to use, Ed was attracted to the simplicity of LoseIt!: “All you can do is log food and weight.”

Streaks, too, is minimalist by design, charging a one-off fee to download. Other apps try to find ways to monetise that serve the overall goal, rather than complicating it. Healthy-eating app Lifesum, for instance, charges users for premium plans that come with specialised diets. Mattias Storm, the company’s head of growth, argues that it provides simplicity for those who need it, with extra support for those who want it. “Habit-forming in general is not trivial,” he says. “It takes 20 days of doing something to get into the swing, and 60 to get into the habit.

“To do it for that long requires a fairly heavy level of motivation. Our users have a very high motivation to live a healthier life … But we’re seeing with some people that they might not want to be tracking what they eat, and so one of the big areas for us right now is around meal plans. You have to understand three concepts to lose weight: how to switch bad foods to healthier meals, portion control and planning ahead. Meal plans are going to help a lot of that.”

In one area of health apps, however, logging a little information can offer useful insights. Ida Tin, the founder of Clue, one of the most popular period trackers on the market, says most apps used to be “calendars that could count to 28 – which is really not that helpful”.

Clue’s processing of inputted data offers an accurate prediction of when your period will begin, which is valuable in a way that passively recording weight fluctuations or cravings may not be. “It’s why we started: we wanted to help women understand their bodies,” says Tin.

Period trackers have another advantage: there is no obvious time to stop using them. Where calorie-counting apps can see users hit their target weight and stop using the app, or gain weight, start using it again, and yoyo painfully for years, with a period tracker, “cycles change. You’re a teenager, you might go through a pregnancy, you might have a period of endometriosis, your body is constantly changing. So you never quite figure it out.

“I think it’s meaningful to have something more like a companion. But one that must, of course, change with you.” Clue even allows users to export a PDF of their data, formatted to be easily understood by a doctor.

That approach has spread from tech to the medical sector. OWise, for instance, allows users undergoing treatment for breast cancer to share information about nausea, sleep quality and fatigue with their doctors. The app “shows benefits for patients and their medical teams”, according to an article in the journal JMIR Cancer.

Healum offers similar tracking for people with type 2 diabetes; My Possible Self is promoted by the NHS as helping to ease stress and anxiety; and Activ8rLives sells smart products, from blood pressure monitors to smart inhalers, that blur the lines between consumer tech and medical devices.

It is a long way from the self-improvement goals of most health apps. But, hopefully, ideas and evidence can flow in both directions: app developers sharing what they know about how to build habits, and medical experts assessing whether those habits really are healthy.

Until then, the best advice is to be driven by your goals, rather than an app; to focus on active change, not passive tracking; and to keep up what works for you. As Ed says: “The main thing that keeps me logging is I don’t want to lose my 310-day streak. Once you’re in the groove, you want to keep going.”

• This article was amended on 12 December 2018 to correct the name of the Streaks app developer to Crunchy Bagel.