State of holidays for subscription apps 2026 report

Every November, LinkedIn fills up with the same advice: discount hard, raise the budget for Black Friday week.

The advice is borrowed from ecommerce.

Black Friday sells discounted coats and TVs, and a meditation app has nothing to put in a cart.

Retail demand spikes that week, and we wanted to know whether subscription demand follows it.

So we counted every US app on Adapty, week by week, October 2025 through January 2026, and marked the five moments the season is planned around: Black Friday, Cyber Monday, Christmas, New Year, and the "New Year, New Me" stretch through mid-January.

Here’s what we found.

The market peaked at +26%. The median app at +5.8%.

Total revenue across all US apps peaked in the first week of January, 26% above a normal October week.

Read that number alone and the season looks like a win for everyone.

Don’t count on the holiday to bump your sales. Instead, try to find your own week.

But if you look at one app at a time, you’ll see that some apps spend November and December below their normal October revenue.

The spread starts widening in November.

By New Year the top tenth of apps sits 163% above its October baseline, the bottom tenth 66% below.

Use your own October as the baseline. The market number will flatter you.

Business peaks on Black Friday. Fitness peaks in early January

Business apps peak in Black Friday week and go quiet by January.

Games peak at New Year, where 36% of Games apps hit their best week.

Health & Fitness and Education bottom out at Christmas, then take early January.

Entertainment holds up in every window.

So a business app puts its weight on Black Friday. A fitness or productivity app works better during the "new year, new me" period. 

Both land their biggest week on January 5-11. Resolutions get made on New Year's Day and paid for the week after.

57% of apps made less on Black Friday than in a normal October week

Plenty of apps treat Black Friday week as the main event. Yet, it's the week apps end up furthest apart.

• 57% of apps made less that week than in a normal October week
• 43% made more, and a third of those made at least 50% more
• Only 18% finished within 10% of a normal week

Add every app together, and the week moved revenue by 0.3%. 

It isn't most apps' big week either. Black Friday was the best week of the season for 7% of them. New Year week was the best for 16%. 

Price cuts don’t explain the gap. Of the apps that made more on Black Friday than in a normal October week, 10% actually lowered their subscription price.

The same was true for apps that made less: 10% cut their price.

Actual price cuts were rare throughout the season. In any given week, fewer than 4% of apps lowered their subscription price, and the cuts around Christmas were just as deep as those around Black Friday.

Whatever separated the two groups, it wasn’t simply a lower price.

Apps showed more paywalls to the same people, and sold less

Across the whole market, paywall views ran 29% above the October average in Black Friday week.

That number is the market total, and it hides two things. 

1. It isn't new people. Unique viewers rose about 6%. The same users saw more paywalls.
2. It isn't most apps. The median app's views ran 5% below October's level. A minority of apps drove the entire increase.

And the extra views converted worse. Conversion per viewer fell to 2.39% the week before Black Friday, against 2.6% in October, and didn't recover until January, when it peaked at 3.34%.

January buyers convert 29% better and refund less

The worry about holiday traffic is that it's cheap attention that churns out in February. The data says otherwise, on every measure we have.

• They convert better. Install to paid was 30.5% for October cohorts and 39.4% for the week of January 12th.
• They refund less. 3.58% in October, 2.91% at Christmas, 3.16% in early January.
• They're worth the same a year later. New Year weekly subscribers are worth $47.54 against October's $48.13, annuals $47.76 against $48.09.

Nothing about your paywall changes in January. The people in front of it are readier to buy, which makes it the traffic worth having your paywall ready for.

New Year's revenue shows up on January 10th

New Year's Day is when people resolve to change something. Trials peak that day, 49% above a normal October day, the busiest trial day of the season.

Nine days later they pay for it. First payments peak on January 10th, 55% above normal, revenue at 38%. The same peak shows up in the weekly view: January 12th to 18th was the busiest week of the season by transaction count, 42% above a normal October week.

Whatever is live on your paywall on January 1st is what that traffic converts against on the 10th.

Weekly plans took share all season, and everything else followed

More buyers picked weekly. It went from 64% of first payments in late October to 71% by the week of January 12th, while annual fell from 21.6% to 18.1% and monthly from 13.2% to 9.3%.

Two things followed from that, and neither was a decision anyone made.

The average payment dropped about 12%. Prices stayed the same, so the drop is the mix: more people buying the cheapest plan.

Intro offers doubled in reach. The cheaper first period covered 12% of purchases in October and almost 25% in January. Nobody switched on a new discount. More buyers just landed on the plans that already had one.

The typical annual subscriber is worth three times the typical weekly one

Year-one revenue per subscriber, net of refunds:

• Weekly: $48.13 average, $12.99 median
• Annual: $48.09 average, $39.99 median

The averages match. The medians don't. Weekly's average is held up by a few subscribers who stay for months, while most pay twice and leave.

The money also arrives at different times. Annual collects 83% of its first-year value in the first 90 days. Weekly collects 60%, and every week in between is another chance to cancel.

What actually moved growth over the holidays

Discounting didn't.

What did work comes in two groups: things you can't change, and one thing you can.

Things that happen to you

• Your category picks your weeks. Business peaks in Black Friday week and is 21% down by January. Games are 36% up at New Year. Health & Fitness is 30% down at Christmas and 13% up in early January.
• Traffic concentrates in a few apps. Paywall views rose 29% across the market and not at all for the median app.
• January buyers are better buyers. Install-to-paid went from 30.5% to 39.4%. Refunds fell from 3.58% to 2.91%.
• The mix shifts to weekly. Weekly went from 64% to 71% of first payments. Introductory pricing doubled, 12% to 24.6%.
• The money lands nine days after the trials. Trials peak January 1, first payments January 10.

The thing you decide: What's on your paywall when they arrive.

And the eight weeks aren't one audience. Black Friday traffic converts worse, January traffic converts better, and the plan people pick changes between them.

Five paywall tips for a successful holiday season

The percentage off is the largest element on most holiday paywalls. It's also the one variable with no signal in our numbers: 10% of apps that beat their October ran a store seasonal offer, and 10% of apps that didn't beat it ran one too.

Five things worth more than the number.

Ship more than one version across the eight weeks

November and January are not the same audience. Conversion per viewer bottoms out at 2.39% the week before Black Friday and peaks at 3.34% in January. Plan mix moves too, weekly from 64% to 71% of first payments. A paywall that runs unchanged from November 24th to January 18th was written for one of those audiences and guesses at the other.

Sell the outcome, not the sale

Every holiday paywall shows a discount. In a week when every competitor's screen says 30% off, the number stops being a reason to pick you.

The screens that work put the outcome first. "Start learning Italian" above the discount. "Get the full Lensa experience for one year" instead of the percentage. A list of what's included, with the saving as a label on one row.

"50% OFF" in the top third of the screen leaves the argument to a caption.

Show the trial timeline, especially in January

Trials peak January 1st, first payments peak January 10th. Those nine days are when people cancel.

A "Today / In 5 days / In 7 days" layout earns its space in January, when most of your revenue is sitting in trials that haven't converted yet.

Put the weight on the plan you want sold

Weekly took share all season, 64% to 71% of first payments. Some of that is your layout: the plan in the biggest type is the plan people pick.

For most apps the plan you want is annual. The typical annual subscriber is worth about three times the typical weekly one, and annual brings in 83% of its first-year money in the first 90 days. Weekly gets to 60% and takes all year.

Match the placement to the week

Black Friday is a one-week moment with a real deadline, which is what full-screen and splash placements are for.

January is a seven-week drift where most of your revenue is sitting in trials that haven't converted.

A splash screen on January 3rd interrupts someone who already started a trial and is deciding whether to keep it. A banner doesn't.

What to do: Interrupt in Black Friday week, stay in the background through January, and put your effort in the trial reminder instead.

Four screens in eight weeks

The season asks for more than one screen. An opener for late November, a gift-led screen for Christmas, a resolution screen for New Year's, and the trial reminder that carries you to January 10th.

Four changes in eight weeks, on a calendar you don't control.

If your paywall ships inside the app binary, every one of those is a build, a review queue, and a staged rollout. Miss the window and you've missed the week.

And the users who never update keep seeing your Christmas screen in February, converting badly, counted in your January numbers.

Your competitors will ship one paywall this season. You can ship four.

Can't design four screens? Prompt them.

Four screens in eight weeks is a design problem before it's a publishing one. Someone still has to draw them.

Or nobody does. Every screen below started as a sentence.

Adapty's skills plug into whichever coding agent you already use, read your live flow, and build a new one from a prompt:

It saves a draft for you to review. It won't publish on its own, and it won't invent a price or a review count.

These took about five minutes.How it works

Your December is now an afternoon.

Gifts, bells, paywalls

The season doesn't lift everyone.

The market total rose, the median app didn't, and at New Year the best tenth of apps sat 163% above its own October while the worst tenth sat 66% below. Revenue moved between apps.

So the market number can't tell you how your season went. 

Discounting didn't decide it. Under 4% of apps ran a store seasonal offer, and the ones that beat their October ran them at the same rate as the ones that didn't.

The week decided it. Black Friday moved the market 0.3% and left 57% of apps below a normal October week.

January is where the season happens: trials peak on the 1st, payments on the 10th, and January buyers convert at 39.4% against October's 30.5%.

That's the best traffic you'll see all year, and it arrives while most teams have stopped looking. Whatever is live on your paywall on January 1st is what converts on the 10th. The coming months are when you have time to build it.

Methodology: how we measured the season

A big app's good week can hide a hundred small apps' bad ones, so every figure compares an app against itself rather than against the market.

Each week is indexed to that app's own normal October week: its October revenue divided by 31, multiplied by 7. A rise means the app beat its own ordinary week.

We report two numbers throughout. The market total adds every app's revenue together. The typical app is the median across apps. In this season those two answers diverge in every cut we ran, which is most of what the report is about.

Week-by-week and app-by-app figures run on a fixed cohort: apps that reported at least 25 days in both October and January, above a minimum daily revenue floor in US October. Without that requirement, apps that left Adapty or had a tracking gap mid-season look like apps that had a bad holiday.

• Every US app on Adapty, October 2025 through January 2026
• 4.7 million first payments and $74.4 million in first-payment revenue
• Peak window means December 28th to January 11th
• A first payment is a subscription starting or a trial converting; revenue is gross, before store commission and tax

Dalibor Vasic