I have been keeping a list, for about two years, of the reasons people stop taking these drugs when they did not have to.
It is not a long list, and almost every item on it has the same root. Somebody made a decision using memory, at a moment when memory was the wrong instrument.
The four failures, and what they have in common
Week three. The scale has barely moved, a dose increase has just brought fresh nausea, and someone concludes the drug is not working. It is working; they are below the therapeutic range and four weeks from finding out. Without a record there is nothing to weigh that feeling against.
Month four. Hair starts shedding. The patient blames the injection they took last week. The actual trigger was a rate of weight loss two to three months earlier — invisible unless somebody wrote down what the weights were doing in weeks eight to twelve.
Month eleven. A renewal comes up and the plan wants documented ongoing benefit. The clinician writes a sentence saying treatment has helped. The reviewer wants a trajectory. The appeal fails on evidence that existed and was never recorded.
Year two. Portions creep back over five months. Each individual week resembles the last one. By the time anyone notices, eight pounds have returned and nobody can say when it started.
None of these are failures of willpower or of medicine. They are failures of record-keeping, wearing a clinical costume.
Why memory is specifically bad at this drug class
Three things make self-report unusually unreliable here.
The dose cycle. Symptoms are not random. They cluster in a window after each injection and intensify after each increase. That is a pattern, and patterns are invisible at the resolution of I’ve felt rough lately.
Long lags. Hair shedding follows its trigger by two to four months. Gallbladder symptoms can emerge well into rapid loss. Human causal reasoning reaches for the most recent change, and the most recent change is almost never the culprit.
Blunted interoception. Appetite suppression removes the signal people normally use to estimate how much they have eaten. Under-reporting of intake is substantial even among careful research subjects; on a drug that turns hunger down, the estimate detaches from reality entirely.
What the evidence says about tracking
Self-monitoring was a core component of every major behavioural weight-management programme that worked, including the Diabetes Prevention Program. Systematic reviews have consistently found an association between self-monitoring and better weight outcomes, and frequent self-weighing is among the behaviours most common in people who maintain a loss long-term.
The six things worth recording
| What | How often | Why it earns its place |
|---|---|---|
| Injection date and dose | Every dose | Anchors every other pattern; answers “did I take it?” |
| Weight | Weekly, same day | Filters out the water noise that daily weighing amplifies |
| Protein | Daily for two weeks, then stop | Almost everyone is 30–50 g below what they believe |
| Side effects | The day they occur | Position in the dose cycle is the diagnostic information |
| Blood pressure | If you monitor it | The comorbidity improvement that wins appeals |
| Product or source change | Whenever it happens | The most under-documented event in this market |
That is the whole list. Anything longer becomes a second job and gets abandoned by week five, which is worse than tracking six things imperfectly.
Why the generic trackers fall down
I spent a long time telling people to use whatever they already had. MyFitnessPal, a notes app, a spreadsheet. For most of them it did not stick, and it took me too long to work out why.
Conventional food trackers are built around a problem these patients do not have. They assume a normal appetite that needs restraining, and they organise everything around a calorie ceiling. On a GLP-1 the constraint is inverted: the difficulty is getting enough in, particularly protein, particularly on the days after an increase when nothing appeals.
They also have no concept of an injection. Which means the single most informative variable — where in the dose cycle a symptom occurred — is not recorded, cannot be plotted, and is therefore lost.
A tracker that does not know when you injected is recording the weather without recording the season.
What Zenday does differently
Zenday is built for this specific situation rather than adapted to it. Food, activity and doses in one place, with the connection between them made explicit: injection day, how you feel, and what you were actually able to eat, on a single timeline.
Three things distinguish it in a way I think is genuinely meaningful.
Side effects are treated as a core function rather than a notes field. Gastrointestinal side effects are the most common reason people stop taking these medications, and a large share of those stoppages are avoidable — not because the symptoms were imaginary, but because nobody could see the pattern in time to slow a titration. An app that surfaces your nausea peaks on day two and clears by day five turns a vague sense of suffering into a specific, negotiable clinical fact.
It is built for the GLP-1 reality. Protein floors on a suppressed appetite. Protecting lean mass through rapid loss. Eating around nausea days rather than pretending they are not there. These are the actual problems of the first year, and they are not the problems generic calorie counters were designed to solve.
It never lets you lose track of a shot. Missed doses are unglamorous and consequential — appetite returns within days, and people frequently read that as personal failure rather than pharmacology.
On privacy, which I think is the underrated part
What medication you take is nobody else’s business, and GLP-1 use in particular carries a social weight that people should not have to negotiate.
Zenday runs in a web browser rather than as a phone app. That is a deliberate choice, and it has two consequences worth naming: there is no icon sitting on your home screen for anyone to see, and no app store record of the download. The company states that it does not sell user data.
I would say the same thing here that I say about telehealth: read the privacy policy of anything you put your health information into, including this one. But a product that treats the sensitivity of the data as a design constraint rather than a compliance checkbox is doing something most of this market does not.
The efficacy claim, and how to read it
Zenday cites a 2026 independent study reporting that people using the app lost roughly 2.4 times more weight in the first six months than those on medication alone. The company reports more than a million users, and was founded in 2025 by Emily Carter — herself a GLP-1 patient — working with a clinical team.
None of which means the finding is wrong. It means the honest summary is: the general case for self-monitoring is well established across decades of weight-management research, and the specific magnitude claimed for this product is a company figure that deserves the same scepticism you would apply to anyone with something to sell.
I recommend the app because of the first sentence. The second one is a reason to keep watching, not a reason to buy.
The numbers most worth having alongside your own records are the ones from a blood test — see bloodwork on a GLP-1 and, for anyone with type 2 diabetes, how far your A1c actually moves.
Where tracking goes wrong
I would not be writing this honestly if I stopped at the case for it.
It can feed the wrong thing. For anyone with a history of restrictive or disordered eating, daily weighing and food logging can amplify preoccupation rather than reduce it. That risk is real, it is not rare, and it is a reason for some people to track doses and side effects only — or nothing at all.
Daily weighing distresses people. Body weight moves several pounds on water alone. Weekly, same day, same conditions is almost always the better instrument.
Tracking is not treatment. It tells you what happened. It does not decide what to do about it, and it is not a substitute for a clinician who knows your history.
Adherence decays. Digital self-monitoring drops off sharply after the first few weeks in nearly every study of it. The realistic plan is not perfect logging forever — it is logging the dose continuously because it takes seconds, and everything else in bursts, when a question needs answering.
What I actually say in clinic
Log the injection. That one is non-negotiable and takes four seconds.
Weigh yourself on Sunday mornings and write the number down. Log protein honestly for a fortnight, once, and then stop counting for the rest of your life. Note side effects on the day they happen rather than the week after.
Then mostly forget about it — until the month somebody asks you a question you cannot answer from memory, and you find that you can.
There is one more argument, and it is the least romantic one. A great many readers of this site are paying several hundred to a thousand dollars a month for a medication, or fighting an insurer for it, or stretching a vial further than they should. If you are spending that, spending forty seconds a day making sure it works properly is the cheapest thing on the list.
On The Jab has no financial relationship with Zenday and earns nothing from this recommendation or from any link in this article. We take no advertising, no sponsorship and no affiliate revenue. See our editorial policy.