Most ecommerce brands do not fail at “having a mobile app for ecommerce” because they picked the wrong button color. They fail because they picked the wrong operating model.
Custom build. No-code builder. Fully managed subscription. Same pitch deck words (iOS, Android, push, fancy screens). Very different cost of ownership after month three.
This is for growth-minded brand owners and ecommerce directors already sitting on heavy mobile traffic. You do not need another feature matrix. You need a decision framework: what each path actually buys you, what it quietly dumps on your team, and when managed is the rational call instead of another tool to babysit.
If the conversion math behind responsive vs native is still fuzzy, start with why your mobile site converts at a fraction of a native app. Then come back here to choose how you ship the channel.
The real decision is ownership, not UI kits
Every serious path can put an icon on the home screen. The split is who owns the unglamorous work after launch:
Store listing copy, screenshots, privacy text, and rejection cycles
OS releases that break login, payments, or push
Catalog and checkout sync when your Shopify, Woo, Presta, or Magento setup is not a demo shop
Push hygiene, segmentation, and creative that does not burn opt-ins
Funnel fixes when mobile CVR stalls two months in
If nobody on payroll treats that as product work, a cheap builder is not cheap. It is deferred engineering with a monthly badge.
Path 1: Custom build (agency or in-house)
You hire a studio or staff mobile engineers. You get a unique UI and total control over edge cases.
When it fits. You already run a product org, you have multi-year mobile roadmap budget, and the app is a strategic surface (marketplace dynamics, heavy B2B rules, custom commerce logic no integrator should touch).
What you actually buy. Code you own, plus a backlog you own. Discovery, design, build, QA, App Store and Play submission, then every OS update forever.
Cost of ownership. Six-figure projects are common once you count product, design, two platforms, and payment edge cases. Timeline often runs a quarter or more before a clean production release. Year two is where the spreadsheet lies: the build team leaves, your roadmap slows, Magento or Shopify plus OS changes still land on you.
Failure mode. Beautiful v1, quiet v2. The app becomes a static brochure while the website keeps getting the merchandising love.
Path 2: DIY and no-code builders
Tools in the Tapcart, Shopney, Plobal-style lane sell speed and a dashboard. Connect the store, theme the screens, publish.
When it fits. Shopify-centered stacks, simple catalogs, and a named owner who will treat the app like a channel every week. Budget wall is real and you accept platform limits.
What you actually buy. Faster first binary. Templates. A plugin surface. You (or an hourly agency) still own QA, submission drama, push strategy, and anything custom past the template.
Cost of ownership. Lower sticker in month one (often low hundreds to around a thousand per month at mid tiers, depending on vendor). Ops time is the hidden line. Every theme oddity, checkout exception, and release becomes an internal ticket. Many of these tools are Shopify-first. Woo, Presta, and Magento operators hit walls faster.
Failure mode. “We have an app” with weak sync, generic UX, and nobody watching activation or push quality. Installs stall. Leadership blames “apps don’t work” when the operating model never showed up.
Semi-managed wrappers sit in the middle: more help than pure DIY, still more of the risk on your side than a full subscription partner. Read vendor scope twice. “We help you publish” is not the same as “we own conversion after publish.”
Path 3: Fully managed on subscription
This is the model ConvertNative runs. Native iOS and Android. Design, deployment, store sync, and ongoing optimisation under a subscription. No multi-year project lock-in theatre.
When it fits. You are an ecommerce operator, not a mobile studio. Mobile is already most of traffic. You want the channel (browse, account, checkout, push, analytics) without staffing iOS and Android. Your stack is Shopify, WooCommerce, PrestaShop, or Magento and you need integration that respects real catalog rules.
What you actually buy. A production app in about 4–6 weeks when scope stays honest, plus a partner still on the hook after launch for sync health, OS churn, and conversion work. Strategy included, not a self-serve editor you abandon.
Cost of ownership. You trade unbounded project risk for a clear monthly operating cost. Your team stays on merchandising, acquisition, and CRM logic. Someone else keeps the binaries store-ready.
Failure mode to watch. Vague scope at kickoff. If checkout exceptions and sacred plugins are not mapped early, any path hurts. Managed does not remove discovery. It removes the part where discovery becomes your permanent night job.
Side-by-side: what directors should compare
Ignore logo walls. Score each path on five boring questions.
Who files the next OS fix? If the answer is “we’ll figure it out,” you chose DIY or a finished agency project.
How does configurable stock and pricing sync? Demo catalogs lie. Your combinations, customer groups, and coupons do not.
Who owns push quality? A 45% open rate channel dies when you blast everyone. See our push notification strategies piece for sequencing discipline.
What happens to the roadmap after go-live? Subscription partners stay measured on outcomes. Project teams stay measured on acceptance criteria.
Can we leave? Data export, subscription terms, and store account ownership matter more than a free trial week.
Write the answers in one page while you shop vendors. Anyone who dodges ownership questions is selling a template, not a channel.
A simple framework you can run this week
Use three filters. Stop at the first hard no.
Filter A: Traffic and leak
Mobile is most of sessions and not most of revenue. Carts die on login, speed, or return visits. Email reach is weak and paid reacquisition keeps climbing. If that is not you, fix site fundamentals first. An app will not save a broken value proposition.
Filter B: Capability
Do you already employ mobile product and eng who ship to both stores on a steady cadence? Yes → custom or heavy DIY can work. No → stop romanticizing “we’ll maintain it in-house” and price managed against the fully loaded cost of one distracted hire.
Filter C: Catalog and stack complexity
Simple Shopify catalog with standard checkout → builders are at least plausible. Deep Magento configurables, messy Woo plugin piles, multi-store Presta → prioritize integration depth and a human owner over a theme picker. Platform-specific notes live in our guides for Shopify, WooCommerce, PrestaShop, and Magento.
If you pass A, fail B, and feel C in your weekly support tickets, managed is the default. Not because builders are evil. Because your constraint is operating capacity, not a missing template.
What “good” looks like when the path is right
Channel quality shows up in commercial numbers, not star ratings alone. Across ConvertNative programs we see about +34% conversion versus the responsive baseline, roughly x2.8 LTV in-app (about $85 vs $30 on mobile web cohorts), around -20% reactivation cost when push carries weight email cannot, and materially higher time spent (up to x10 in retained sessions). Push open rates around 45% (near 7x typical email) only matter if checkout and catalog sync hold.
Scale reference: Notino at 7M downloads and retail volume. Your catalog is not Notino. The lesson still holds. Native pays when browse, account, checkout, and messaging stay maintained after launch day.
For the retention math CRM leaders care about, keep push open rates vs email next to this decision doc. For the weekly metrics habit after you pick a path, use ecommerce app analytics that change weekly decisions.
Common objections, answered without spin
“We’ll start with a PWA.” Fine as a patch. Weak as a lock-screen channel. You still lack reliable push economics and home-screen habit on both major mobile platforms the way a native app delivers.
“Builders are enough until we scale.” Sometimes. If you are already scaled on traffic and leaking on mobile CVR, “until we scale” is how another year passes.
“Managed means less control.” You keep merchandising, offers, and brand rules. You delegate binary maintenance and optimisation labor. Control over revenue levers matters more than control over Xcode.
“We tried an app once.” Ask whether you tried a channel with an owner, or a project with a launch party. Most scars come from the second.
Pick the path that matches the job
Build custom when mobile is a core product surface and you staff it like one. Buy a builder when the catalog is simple, the platform fits, and a named owner will live in the dashboard. Go managed when mobile revenue is the job and running two app stores is not.
If you want a clear read on which bucket you are in, bring mobile share of traffic, mobile CVR vs desktop, and your stack constraints. We will say if a native app is worth it this quarter, and whether managed is the right shape.
Get a free mobile app audit. Straight scope, no feature theatre.