What web shop conversion rate should you actually expect

Spotify moved first after the 30 April contempt ruling, then Patreon, then Kindle, all of them pointing US iOS users toward web checkout before a single mobile game publisher had shipped a link-out flow of its own. Game studios assembling their first web shop now want to know what conversion rate to build a forecast around. The honest answer is unhelpful at first: the early non-gaming examples won't tell them.

What "conversion rate" actually means here

Two steps, not one. Conflate them and the forecast goes wrong in both directions at once, because the first step is the click-through rate on the external link itself, out of everyone who saw it inside the app, while the second is the purchase conversion rate on the web page that link leads to, out of everyone who actually clicked. A studio tracking only the second number will look far more efficient than it is. It has quietly dropped everyone who saw the link and then did nothing.

Why the early non-gaming examples do not transfer

Spotify and Patreon sell subscriptions to services the user already understands and has usually decided to buy before opening the app at all. A game purchase works differently. It tends to be an impulse inside a session, prompted by something specific on screen: an event timer, or a currency shortfall at exactly the wrong moment. Send that impulse out to a browser and you have introduced a context switch the subscription purchase never had to survive, and every extra step between the moment of intent and the moment of payment costs conversion. E-commerce has known this for years; each additional checkout page a customer must click through takes a meaningful share of completions with it, and no mobile game's purchase intent is likely to prove more durable than an abandoned cart.

A worked example with hypothetical numbers

Take a hypothetical mid-size puzzle game with 100,000 daily active users, where the in-app store historically converts 3% of daily actives into a purchase on a given day. The studio adds a web shop CTA visible to all 100,000. A reasonable planning assumption, and to be clear that is a planning assumption rather than an observed industry figure, might run like this: 15% of the users who would otherwise have purchased in-app click the web shop link instead, because the other 85% either buy in-app anyway where the rules still permit it, or never notice the new option at all, or simply aren't in a purchase-intent moment when the CTA appears. Then comes the drop-off between click and completed payment, plausibly a third to a half of clicks, given the friction of leaving the app and typing card details into a browser. Plan on that basis and the web shop channel takes a minority share of its addressable purchase intent over its first months live. Not parity with in-app conversion, and certainly not better.

The common early mistake is to treat the web shop CTA as a design problem. Get the button colour right, get the copy right, get the placement right: conversion follows. It doesn't. Placement relative to purchase intent matters considerably more than any of that, because a CTA sitting permanently in a settings menu converts at a fraction of the rate of one that surfaces at the exact moment a player runs short of the currency needed to finish a purchase they were already about to make in-app. The studios most likely to beat the conservative range above treat CTA timing as the primary lever and visual design as the secondary one, rather than the reverse.

Setting expectations before the QBR

None of this arithmetic produces a precise forecast, and it isn't meant to. It exists to stop a growth lead walking into a quarterly review with a projection lifted from a Spotify case study, then explaining the shortfall as an execution failure rather than a mismatch of comparison. As covered in "The Web Shop Attribution Blind Spot Nobody Budgets For," most of the measurement gap on this channel comes from what a team fails to track, not from conversion behaviour that is unusually poor.

So: build the forecast on the two-step framework above, track both steps separately from week one, then treat the first quarter of real data as the actual benchmark rather than anything published by a company selling a different kind of purchase.

What a healthy trajectory looks like over time

Both steps improve with iteration, which makes month one the wrong thing to argue about. The better planning question is how fast the numbers should move afterward. A studio measuring click-through and purchase conversion separately from launch should see the first improve gradually as CTA placement gets refined against real purchase-intent moments, and the second improve as friction comes out of the web checkout flow itself. Two or three iteration cycles with neither number moving is not a slow-conversion forecast; it points to a targeting or friction problem, and that wants direct investigation rather than more patience.

What improves the number once it exists

Once real click-through and purchase-conversion data exists, the effective levers are the ones e-commerce checkout teams have used for years: cut the number of taps between the CTA and a completed purchase, pre-fill whatever the platform allows, and time the CTA's appearance to a genuine in-game purchase-intent moment instead of leaving it up permanently. None of that waits on a platform changing the rules again. It asks a team to treat the web shop as its own funnel with its own optimisation backlog, rather than a compliance checkbox added once and left alone.

Related archive reading

These articles provide related context and remain subject to their stated review status.

• The Web Shop Attribution Blind Spot Nobody Budgets For
• Web-to-app funnels for mobile games