Writing Nutrition
Why nutrition apps lose most of their users by day 30
The retention problem in food tracking is not a feature problem. It is the failure state built into the design, and removing it changes what the product has to be.
Health and fitness apps typically retain somewhere in the region of one in ten users to day 30; the strongest performers reach roughly a quarter. Everyone in the category knows these numbers. The usual response is to add features (recipes, challenges, badges, a social feed) on the theory that a stickier app holds more people.
We think the numbers describe something else. People do not leave because the app lacked a recipe library. They leave because the app was built in a shape that requires them to be perfect.
The shape of a diet
Almost every tracker, whatever its branding, has the same three structural elements:
- A number you must hit. A calorie target, a macro split, a goal weight.
- A plan to follow. Explicitly, as meal plans, or implicitly, as the eating pattern that hits the number.
- A visible failure state. The day you go over, the streak you break, the red bar, the “you’re 620 calories over” summary.
Element three is not a side effect; it is the engagement mechanic. Guilt is a strong short-term motivator, which is why the design keeps getting rediscovered. It is also a terrible long-term one, because the first genuinely disrupted day, whether a birthday, a work trip or a sick child, produces a failure, and failure produces avoidance, and avoidance is a deleted app.
Ask anyone who has quit a tracking app when they stopped. Almost nobody says “when I got bored.” They say “when I went on holiday”, “when I had a bad week”, “when I couldn’t be bothered to log the third day in a row and it felt pointless to start again.”
What “no failure state” actually requires
It is easy to say a product is not judgemental. It is harder to build one where there is genuinely nothing to fail, because the constraint removes tools most of the category depends on.
No target as the default. If the home screen leads with a number to hit, everything else is decoration. Awareness (here is what you ate, here is the pattern over a week) is a different screen and a different product.
Streaks that reward logging, not outcomes. The habit worth reinforcing is the record-keeping, because that is the thing the user actually controls. And a missed day should freeze the streak automatically rather than reset it. A streak that punishes an ordinary human week is a countdown to churn.
Copy with no moral vocabulary. No “good” or “bad” foods, no “you blew it”, no cheerful passive aggression about yesterday. This sounds cosmetic and is not: tone is where diet culture re-enters a product that claims to have removed it.
Friction low enough for a bad day. If logging takes ninety seconds of searching a database, it will not survive a hard week, and hard weeks are exactly when the record matters most. Photo and voice input exist for that reason, not for the demo.
The trade we are accepting
This is not a free lunch, and it would be dishonest to present it as one.
Removing the failure state removes the sharpest short-term engagement lever in the category. A guilt-driven app can show better week-two numbers than we will. Our bet is that day-30 and day-90 look different, because the reason people quit is gone.
That is a testable claim, not a slogan, and it is the one number we hold ourselves to. If retention past day 30 does not beat the diet-shaped alternatives, the thesis is wrong and no amount of positioning saves it.
What this means if you are just looking for an app
Some practical questions worth asking of anything you are considering, ours included:
- What happens the day you go over? Look at the actual screen, not the marketing.
- What breaks your streak, and what does the app say when it does?
- Can you log a restaurant meal you already ate, in one step, from memory?
- Where does your data go? “We never sell your health data” should be a written commitment, not an implication.
- Can you delete your account and your history from inside the app?
Those five answers tell you more about how a food app will treat you in month three than any feature list.
Our take on all five: the nutrition app · Trust & data