Open your banking app and scroll back through the last three months. Count every charge that shows up on the same day each month or each year. Streaming, music, cloud storage, a fitness tracker, a language app you opened twice, maybe a meal planner your sister talked you into. Most people land somewhere between eight and fifteen, and at least two of them surprise the person doing the counting. Check the annual ones too. A $59.99 renewal in October is easy to miss when it sits between a grocery run and a gas fill-up.
That list looks like a trap, and sometimes it is. But subscriptions took over apps because the math works for both sides of the screen: the maker gets revenue that arrives every month instead of once, and the user pays a small amount up front for something that keeps getting fixed and updated. The rest of this article tests five things people say about that arrangement against the numbers and against what I have seen shipping these products for clients.
“Subscriptions are just a way to charge you forever”
What people say. An app that used to cost $4.99 now costs $4.99 every month, and the only thing that changed is the greed of the people who make it. The software did not get ten times better, and the price tag now sits on your statement twelve times a year instead of once on a receipt you forgot about.
What actually happens. The platforms pushed makers this way first. For years Apple and Google kept 30 percent of every sale, which meant a $2.99 app left its developer with about $2.09 and no way to fund the next year of work. Apple’s App Store Small Business Program started on January 1, 2021, and cut the commission to 15 percent for developers earning under $1 million a year. Look at it from a small studio’s side. A team of three, earning maybe $400,000 a year, suddenly keeps 85 cents of every dollar, and the only way to turn that into a salary next year is a customer who pays again next month.
Then the money followed. Sensor Tower’s annual report, published January 21, 2026, put global in-app purchase revenue at $167 billion for 2025, up 10 percent, and for the first time non-game apps brought in more than games. The report does not break that out by billing type. My guess, not their figure, is that subscriptions drive most of it. What the report does show is that the old arithmetic of selling a million copies and moving on has given way to a business that has to keep earning its place. A charge that repeats is the price of that change, and it is also the reason the app still gets patched in year four.
“Paying once was fairer”
What people say. Buy it, own it, done. Nobody sends you a bill for a calculator, and a lawn mower does not come with a monthly fee for owning it.
What actually happens. A calculator is a good example of where that works, and almost nothing on your phone is a calculator anymore. Last year a founder came to us with a meal-planning app and wanted a single $29.99 price, the way software was sold when he was in college. We walked through what the app does: it pulls recipes from a server, syncs the grocery list between two phones, calls a nutrition database, and stores years of saved plans. Every one of those has a monthly bill that keeps arriving whether or not the customer ever pays again. He wanted to cover an unknown number of years of hosting with one payment.
Founders usually reach this point in the first week of working with a team that offers SaaS application development services, and the plan structure ends up being the first product decision rather than the last. He went with a monthly plan and a discounted annual one. A one-time price works fine for apps that live entirely on your device, such as a unit converter or an offline puzzle game, and I will say so to any client who builds one. For anything that talks to a server, the one-time price quietly assumes the customer leaves.
There is also the shelf-life problem that nobody remembers fondly. The ugly truth about the old model is that the developer usually lost interest before the customer did. Plenty of paid-once apps from a decade ago stopped launching after an iOS update because the developer had already moved on. The buyer got ownership of something that no longer opened. I have a folder on an old tablet full of those, and I paid for every one.
“Everyone is exhausted by subscriptions, so people will stop”
What people say. Surveys keep showing that consumers are fed up. The model has to crack soon.
What actually happens. The fatigue is real, and the data is blunt about it. Deloitte’s Digital Media Monitor, updated October 15, 2026, found that subscribing households spend an average of $76 a month on streaming video, almost two-thirds of consumers are frustrated that prices keep rising, and 43 percent canceled a paid streaming service in the previous six months. Back in 2022, Bango asked 2,500 US subscribers, and 72 percent said there were already too many subscription services.
Notice what those numbers describe, though. People are canceling, and they are also still paying $76 a month. They are not leaving the model. They are auditing it. At home, my wife and I sit down every few months with the bank app and go through the recurring charges one by one. The last round killed a language app I had used twice in a quarter and kept a three-dollar weather app that I open every morning. Nobody forced either decision, and the apps that stayed were the ones I would have defended in an argument.
From the maker’s side this is brutal and also healthy. One logistics client of ours lost roughly a fifth of its paying users over a single winter to what the industry calls churn, meaning subscribers who cancel or lapse. When we asked the ones who left why, the usual answer was some version of “I still liked it, I just wasn’t using it.” Liking an app stopped being enough. I think fatigue works as a filter: fewer apps survive each household’s review, and the ones that do tend to be tied to a daily habit. That filter is my interpretation. The surveys only measure the frustration. What I can say from watching renewal dates is that the apps that hold on to people share one trait: they are in somebody’s week. A fitness app attached to a Tuesday run survives. A planner that generates Sunday’s grocery list without being asked survives. An app that waits to be remembered does not, and a monthly charge is a very good reminder of that.
“A long free trial just gives the product away”
What people say. Three days is plenty to judge an app. Anything longer and people take what they need and leave.
What actually happens. Most apps run short trials, and the data says that is the wrong call. RevenueCat’s State of Subscription Apps 2026 covers more than 115,000 apps and over $16 billion in revenue. In it, trials of 17 to 32 days convert at a median of 42.5 percent, while trials of four days or less convert at 25.5 percent. Even so, 46.5 percent of apps now run the short kind, up from 42.1 percent a year earlier.
Two other numbers from the same report stuck with me. Apps that put up a hard paywall (pay or leave, no free version) converted trial users at a median of 10.7 percent by day 35, against 2.1 percent for freemium apps, where the basic version is free and you pay to upgrade. And 35 percent of all annual-plan cancellations happen in the first month. I read that as people grabbing the yearly discount before they had formed the habit, then realizing they had paid for a year of something they did not use.
We saw the same shape on a wellness app that moved from a three-day trial to two weeks. Nothing about the product changed. People simply had time to do the thing twice, which is what makes a $9 charge feel reasonable. A short trial is a bet that the first impression is enough, and the report suggests it usually is not.
The fear behind the objection is fair, because some people do take what they need and leave. In my experience that group is smaller than founders expect, and it is not the group paying for the product anyway. The users who matter are the ones who need a few weeks to find out whether the app fits their routine. Cutting the trial to three days mostly cuts them off. What I would add is a reminder before the first charge, not after, since a surprise bill is the fastest way to turn a happy trial user into a one-star review.
“If a payment fails, that’s the customer’s problem”
What people say. The card expired, so the person lost access. Their fault, their loss.
What actually happens. This one costs makers real money, and most of them do not notice because it hides inside the churn number. In the same RevenueCat report, billing failures cause 31 percent of cancellations on Google Play and 14 percent on the App Store. Think about how that feels from the customer’s chair. You wake up one morning, the app says your plan has lapsed, and you never made a decision. Those are not people deciding to leave. Those are cards that expired, banks that declined a charge they did not recognize, and balances that were a few dollars short on the wrong day.
Vladyslav Kuzniak builds the backend for several of our subscription products, and he has dealt with this more than anyone on our team.
“On one Android app we were losing about one in four subscribers to failed card charges, and the founder read it as churn. We added three retries spread over 16 days, a grace period that kept the account open, and one in-app banner saying the card had failed. We recovered roughly 18 percent of what had been written off as cancellations. The code took four days. Getting the team to treat billing as a product feature instead of plumbing took longer.” (Vladyslav Kuzniak, Backend Developer at LITSLINK)
The industry calls this retry-and-notify routine dunning, which is an old word for chasing a debt and a bad fit for what is mostly a friendly reminder. The fix is cheap compared to winning a new customer, which is why I get a little impatient when a team spends a quarter on ad campaigns and a day on retries. Both numbers belong on the same dashboard. Whoever watches signups should also be watching the share of renewals that quietly fail. What I tell clients is that nobody owns the failed-payment emails by default, so somebody has to be given them in the first sprint, not the last. When billing gets built in the final week, the paywall shows up at a random moment, the trial length is whatever the store defaulted to, and the person whose card expired in March gets an email that reads like a collections notice.
Here is where I think the subscription model deserves its worst reviews. Hiding the cancel button is a real practice, and clients have asked us to do it. We push back. Apple and Google both route cancellation through their own subscription settings, so burying it inside the app mostly annoys people before they find the official route. On one wellness app we moved cancellation to a single tap, and cancellations did not rise, while support tickets about it dropped. I would not call that a controlled test. It was one app over a few months, and I am reporting it as what we saw. The same goes for price increases. A hike announced in a newsletter nobody reads looks like a sneaky charge on the next statement, while the same hike shown in the app with a month of notice and one honest sentence about why tends to lose fewer people. We never ran that as an experiment either, but two clients showed us the same pattern, and we stopped doing it the first way.
Before you renew
Next time a renewal email lands, put three questions to the app. When did I last open it, and did I open it because I wanted to or because a reminder pushed me? If this charge doubled tomorrow, would I pay it, or would I be relieved to have an excuse to cancel? And can I find the cancel button in under a minute without contacting anyone? If the answers are last week, yes, and yes, keep paying. Anything else is a vote you are allowed to change, and the app should be glad you were asked.
