The mobile app pitch is the most persuasive deck in DTC right now, and it is persuasive because none of it is made up. Ecommerce apps do convert at 2 to 6 percent against 1.5 to 3 percent on mobile web. App customers do show lifetime values several multiples above web-only buyers. Published benchmarks do put app revenue at 20 to 60 percent of online sales, generated by under 15 percent of users.
Then you look at adoption. Mobile app penetration across Shopify Plus stores sits under 2 percent. If the returns were as clean as the deck suggests, that number would not be under 2 percent. Something in the gap between the published lift and the actual adoption is worth understanding before you sign a twelve month contract.
The numbers are real. The comparison is not.
Think about who downloads a brand app. Almost nobody installs an app from a company they have never bought from. The install happens after a good first order, usually after a second, usually from someone who already had the brand in their head. By the time a customer is in your app, they have self-selected into the most loyal decile of your base.
So when a platform reports that app users convert 3x better and carry 2.8 to 5x the lifetime value, it is comparing your best cohort against your entire traffic mix. Some of that lift is the app. A large share of it is the fact that these people were always going to buy again. The app captured the behaviour. It did not create it.
The same logic applies to the revenue share figure. An app that takes 25 percent of your online revenue has not added 25 percent to your business. Most of those orders moved across from email, SMS, paid retargeting and direct. The question is never what share the app takes. It is how many of those orders would not have happened at all.
An app does not manufacture loyalty. It gives loyalty somewhere better to go. If you do not have loyalty yet, you are paying a monthly fee to house an empty room.
The four conditions that have to be true first
I have run this assessment for supplement, skincare, food and drink and apparel brands, and the answer is the same every time. An app pays when all four of these hold. It loses money when even one is missing, because the fixed cost does not care how few people installed.
Purchase frequency above three orders per year
This is the gate that disqualifies most brands. An app is a retention surface. Retention surfaces only earn their keep when there is something to retain to. If your customer buys once a year, a home screen icon is a monthly reminder of a decision they are not ready to make again.
Consumables clear this comfortably. Supplements, coffee, skincare refills, pet food, cleaning products, anything with a natural replenishment cycle under four months. Repeat purchase rates in those categories run 40 to 55 percent, against 25 to 32 percent in fashion and around 10 percent in luxury. If you sit at the bottom of that range, the app economics will not rescue you.
A repeat base large enough to produce a real cohort
Install rates among existing customers realistically land between 10 and 20 percent, and that is with active promotion across email, SMS, packaging inserts and a first-order incentive. Run that against your actual repeat customer count rather than your total list.
A brand with 40,000 customers and a 28 percent repeat rate has roughly 11,200 repeat buyers. A 15 percent install rate gives you about 1,700 app users. If those users each place two extra orders a year at a 60 GBP average order value and a 40 percent contribution margin, the app is contributing roughly 82,000 GBP in gross profit before you subtract platform fees and internal time. That works. Halve the customer base and it does not.
Email and SMS already running properly
Push is an owned channel, and owned channels compound only when they are run with discipline. Automated push sequences make up roughly 5 percent of push sends in ecommerce but drive 28 percent of push-attributed orders, which is the same pattern you see in email: the triggered flows carry the revenue and the broadcasts mostly create fatigue.
If you have not built that discipline in Klaviyo yet, you will not suddenly find it in a new channel with a more intrusive delivery mechanism. Brands that arrive at push with weak lifecycle habits blast promotions, drive uninstalls in the first fortnight, and end the quarter with an app nobody opens. Get the flows right where the stakes are lower.
Average order value that covers the running cost
App subscriptions across a full Shopify stack typically run 0.5 to 1.5 percent of net revenue, and a dedicated mobile app platform sits at the upper end. That is the visible number. The invisible one is operational: merchandising the app, building the push calendar, keeping product data and creative in sync, and maintaining store listings.
Work out how many incremental orders per month you need to cover the platform fee plus those hours, then compare it to your realistic install base. At a 30 GBP average order value and thin margins, the maths is brutal. At 80 GBP and above with healthy contribution margin, it gets comfortable quickly.
What to fix instead, in the order that pays
The average DTC brand retains 28.2 percent of customers for a second purchase. One study of 156,110 customers earlier this year put the aggregate repeat purchase rate at 18.8 percent. Nearly three in four first-time buyers never come back, and roughly 60 percent of DTC revenue comes from the ones who do. That is where the money is, and none of it requires a new platform.
Start with the second purchase window. A triggered, product-specific re-buy invitation sent inside the first 30 days lifts repeat purchase rate by 8 to 14 points within a quarter for most brands. Not a generic newsletter. A message that names the product they bought, the moment they are likely to run low, and the single next thing worth adding. Most brands send a review request in that window and nothing else.
Then fix replenishment timing. Pull the actual median days between first and second order per SKU rather than guessing at 30 or 60. In most catalogues I audit, that number differs by two to three weeks across the top five products, and every flow is firing on the same schedule. Aligning the trigger to real consumption data is a one afternoon change that moves the curve.
After that, win-back sequencing and post-purchase education. Only once those are running properly does the app question become interesting, because now you have a repeat base worth installing something against.
If you do build one, run it like this
Assume push behaves like email, not like a megaphone. Build the triggered sequences first: replenishment reminders, back in stock, price drop on a saved item, order status. Keep broadcast pushes to one or two a week maximum and give people category-level control over what they receive. Uninstalls are permanent in a way that email unsubscribes are not, because reinstalling requires a trip to the app store and a reason.
Give the app a reason to exist beyond convenience. Early access to drops, an app-only bundle, loyalty balance and reorder in two taps. If the app is just your website with a home screen icon, install rates stall around 5 percent and the cohort never gets big enough to matter.
And measure it honestly. Match installers to a control group with similar order history, recency and category mix, then compare orders per customer over 90 and 180 days. That difference is your real lift. The platform dashboard will credit the app with every order placed inside it, which tells you where the order happened, not whether it would have happened anyway. Those are completely different questions and only one of them affects your P&L.
What this looks like in practice
A supplement brand I work with came to me with an app quote and a launch date. Their repeat purchase rate was 21 percent against a category benchmark in the forties. Their replenishment flow fired at day 45 for every SKU, while their actual median reorder gap ranged from 26 to 61 days depending on the product.
We parked the app and spent the quarter on the boring work. Per-SKU replenishment triggers, a 14 day post-purchase education sequence built from their own review language, and a win-back offer segmented by original product rather than by recency alone. Repeat purchase rate moved to 31 percent. Email revenue share went from 19 to 28 percent.
They are building the app now. The difference is that they are launching it against a repeat base that is roughly 50 percent larger than it was, with lifecycle discipline already proven in a cheaper channel. Same app, completely different economics, because the sequencing was right.
Inside the system
How we build this for brands
The retention work above is not manual in our stack. We run a VOC engine that mines reviews and support messages to find the language customers actually use about running out, switching, or staying, and that language becomes the post-purchase education sequence rather than something a copywriter invents. Alongside it, lifecycle flows get built and deployed into Klaviyo by AI against real per-SKU reorder data, so replenishment triggers land on consumption timing instead of a round number someone picked.
A reporting agent then watches repeat purchase rate, second order window and contribution margin weekly and surfaces the leak before it costs a quarter, which is also what tells us whether a channel like an app has earned its place yet. Part of this runs live for portfolio brands today; the full system is what we deploy when we take a brand on.
Retention Audit
Find Out Whether An App Would Actually Pay
I will run your repeat purchase rate, second order window and per-SKU reorder timing against category benchmarks, size the realistic install cohort, and tell you whether an app clears the bar or whether the money belongs in your flows this quarter. You get the numbers and the decision, either way.
Book Your AuditFrequently asked questions
Is a mobile app worth it for a DTC brand?
Only when four things are already true: customers buy three or more times a year, your repeat base is large enough that a realistic 10 to 20 percent install rate produces a meaningful cohort, your email and SMS programme is already mature, and your average order value covers the platform fee plus the internal time to merchandise it. Fewer than 2 percent of Shopify Plus stores run a mobile app, which tells you it is a niche play rather than a default. If your repeat purchase rate is under 25 percent, the app is not the bottleneck.
Do ecommerce mobile apps really convert better than mobile web?
Yes, but the gap is smaller than it looks once you account for who installs. App conversion typically runs 2 to 6 percent against 1.5 to 3 percent on mobile web. The people who download a brand app are almost always existing customers who already intended to buy again, so you are measuring your best cohort against your whole traffic mix. Treat the headline multiple as an upper bound and test incrementally.
What percentage of ecommerce revenue comes from a brand mobile app?
Published figures put app revenue at 20 to 60 percent of online revenue, usually from under 15 percent of users, with individual brands reporting numbers around 20 percent of total revenue. The useful question is not what share the app takes but how much of it is genuinely incremental rather than orders that would have arrived through email, SMS or direct anyway.
How do I measure whether a mobile app is incremental?
Compare cohorts, not channels. Match installers to a control group with similar order history, recency and category mix who did not install, then measure orders per customer over the following 90 and 180 days. The difference is your honest lift. Platform dashboards attribute every app order to the app, which tells you where the order happened, not whether it would have happened anyway. If you cannot build a matched cohort, hold a region or a segment out of install promotion for a quarter.
What should DTC brands fix before building a mobile app?
The second purchase window. The average DTC brand retains 28.2 percent of customers for a second order, with category benchmarks running from around 10 percent in luxury to 40 to 55 percent in consumables. A triggered, product-specific re-buy invitation inside the first 30 days lifts repeat purchase rate by 8 to 14 points in a quarter for most brands. That is a flow build, not a platform build. Fix replenishment timing, post-purchase education and win-back sequencing first.
How much does a Shopify mobile app cost to run?
The platform fee is the smaller number. App subscriptions across a Shopify stack typically run 0.5 to 1.5 percent of net revenue and a dedicated mobile app sits at the higher end of that range. The real cost is operational: merchandising the app, building the push calendar, keeping product data and creative in sync, and maintaining app store listings. Budget the internal hours before you sign, because an unmerchandised app decays fast.
About the author
Caner Veli is a DTC operator who has helped 350+ brands fix broken growth engines. He built Liquiproof from zero to 3,000+ global retailers in under 6 years. He now runs the same playbook, supported by AI systems he built himself, for DTC and CPG brands.