Freelancer, no-code or agency: who should build your app?
10 min read
Key takeaways
- The mobile site acquires, the app retains: they are not competing options but two floors of the same funnel.
- An e-commerce app becomes profitable when a significant share of your revenue comes from customers who buy several times a year.
- A PWA covers 80% of the need for 30 to 50% of the price, but stays weak on discovery, installation and home-screen presence.
- Push notifications only pay off if they are rare, segmented and useful; a daily promo gets the app uninstalled.
- Budget for an iOS + Android app connected to your catalogue, plus monthly maintenance: free quote within 24 hours, at low prices.
The question almost always arrives in this form: "our competitors have an app, do we need one?" It is the wrong question. An e-commerce app (in French) does not replace your mobile site, it does not bring in new customers, and it is expensive to maintain. On the other hand, well positioned, it can become the most profitable channel of your shop.
The right question is therefore: do you have enough returning customers for a channel dedicated to their loyalty to earn more than it costs? This article gives the method for answering it, with the real costs of each option: optimised mobile site, PWA, native app.
Three options, three different roles
Before comparing prices, you need to understand that these three solutions do not do the same job.
| Optimised mobile site | PWA | Native app | |
|---|---|---|---|
| Main role | Acquire: SEO, advertising, social media | Make returning easier without going through the stores | Retain and drive repurchase |
| Discovery | Google, shared links | Google, then an invitation to install | Stores, but mostly from your site |
| Installation | None | Manual, not very intuitive on iPhone | From the App Store or Google Play |
| Push notifications | No (or limited web push) | Yes, with restrictions on iOS | Yes, complete and reliable |
| Fast payment | Apple Pay / Google Pay possible | Same as the site | Same, plus an always-open session |
| Hardware access | Limited | Limited | Full: camera, NFC, background geolocation |
| Indicative initial budget | Included in the site | Layer on top of the site, on quote | On quote |
The key point is in the first row. The mobile site is your front door, and it must stay excellent whatever happens: nobody downloads an app to discover a brand. The app comes afterwards, for those who have already bought and will buy again.
The deciding rule: if your mobile site is slow, poorly designed or converts badly, do not invest a single euro in an app. Fix the site first. An app never makes up for a failing checkout funnel: it inherits its flaws, with one more budget to maintain.
Why an app can earn more per customer
Users of an e-commerce app generally buy more often than website visitors. It is not magic, and you need to understand why so as not to be wrong about expectations.
1. A self-selection bias, openly acknowledged. Those who install the app are already your most engaged customers. Part of the performance gap comes simply from that: they would have bought anyway. The app does not create this loyalty, it exploits it better.
2. Friction disappears. Session always open, address and payment method remembered, basket kept between two sessions, payment in one gesture. On a website, each return may require logging in again, a forgotten password, re-entering details. Every step removed wins back purchases.
3. Presence on the home screen. Your icon is seen several times a day. It is a free reminder that the website cannot obtain without paying for advertising.
4. A direct channel that depends on nobody. The push notification reaches your customer without going through a social network algorithm or a spam filter. It is the only retargeting channel whose access you do not rent.
5. Uses that are impossible on the web. Scanning a barcode in store, retrieving receipts, presenting a loyalty card, being alerted when approaching a shop. This is where physical retailers often find their best argument.
The profitability calculation, in practice
An e-commerce app is justified by the additional purchases it generates, not by its number of downloads. The reasoning comes down to four questions.
How many repeat customers do you have? Count those who have bought at least twice in the last twelve months. That is your real potential audience, not your total traffic.
What share will install the app? Only a fraction, even with good promotion on the site, in emails and at the checkout. Be cautious in your assumptions: reason on a minority of your repeat customers.
How many extra purchases per year? Not the total revenue from these customers, only what they will buy in addition thanks to the app.
What margin on these purchases? It is the margin, not the revenue, that must cover the investment.
Let us take a theoretical example to illustrate the mechanics. Add up the development cost and twelve months of maintenance: that is the first-year investment, after which only the maintenance recurs each year. If your gross margin is 40%, you need to generate two and a half times that amount in additional revenue to break even, in the first year and in the following ones.
2 purchases / yearFrequency threshold below which an app is unlikely to pay for itself
Low priceBudget for an iOS + Android e-commerce app connected to an existing catalogue, free quote within 24 hours
Fixed feeMonthly maintenance, iOS and Android updates included
This calculation brings out clear profiles.
| Shop profile | Verdict |
|---|---|
| Repeat purchases (food, beauty, sport, fashion, pets) and a large customer base | Native app often profitable |
| Retailer with physical stores and a loyalty programme | Native app highly relevant: card, scan, click & collect |
| DTC brand with an active community and drops or limited sales | App relevant, especially for notifications |
| Rare or high-value purchase (furniture, appliances, travel) | Excellent mobile site, possibly a PWA |
| Catalogue under construction, fewer than a few thousand customers | Mobile site first, app later |
The PWA: the honest middle ground
The Progressive Web App is often presented as "an app without the drawbacks". That is partly true. It reuses your site, installs on the home screen, works partly offline and can send notifications. A single codebase, no store validation, instant updates.
Its limits are nonetheless real, and they affect precisely what gives an e-commerce app its value.
Installation remains marginal. On iPhone, you have to go through the share menu and then "Add to Home Screen". Few visitors do it without being guided, and you cannot trigger a native install prompt as on Android.
Notifications on iOS require installation. Web push has been supported on iPhone since iOS 16.4, but only for a PWA added to the home screen. No installation, no push. Your subscriber base will therefore be clearly smaller than with a native app.
No presence in the stores. Some customers search for your brand directly in the App Store. A PWA is not there.
More limited hardware access. Barcode scanning works through the camera, but NFC, background geolocation or integration with wallets are more constrained.
In practice, the PWA is an excellent choice for a shop that wants to improve the experience of its mobile customers, test the appetite for an installed channel, or whose budget does not yet allow a native app. It remains the most economical option if the existing site is technically sound, and clearly more expensive if it has to be reworked first.
Push notifications: the lever that can also destroy
Notifications are the main reason to have an app. They are also the main reason people uninstall it.
What works:
- Transactional notifications: order confirmed, shipped, delivered, ready in store. Useful, expected, never perceived as intrusive.
- Alerts explicitly requested: an item back in stock, a price drop on a favourite, a private sale opening. The customer said yes to this specific information.
- The abandoned basket, once, with a reasonable delay.
- Segmented content: a new arrival in the category the customer actually buys, not across the whole catalogue.
What gets the app uninstalled: the generic daily promotion, notifications sent to the whole base at the same time, and anything that looks like a newsletter pushed onto the lock screen.
A simple rule: every notification must be justifiable by an action or a preference of the customer. If you cannot say why this person receives this message, do not send it. And ask for permission at the right moment, after a first purchase or on a stock alert, rather than at first launch, when the user has no reason yet to say yes.
On compliance, marketing notifications fall under the same consent principles as the rest of your communication: the user must be able to refuse them and unsubscribe easily. Our guide on GDPR compliance for a mobile app details the points to check.
What an e-commerce app really costs
An e-commerce app is not a shop rebuilt from scratch. It plugs into your catalogue, your stock management and your existing back office (Shopify, PrestaShop, WooCommerce or a custom platform). The budget depends mostly on what the app does beyond the site.
| Item | Share of budget | What makes it vary |
|---|---|---|
| Scoping and design | 15 to 20% | Number of screens, existing design system or not |
| Catalogue, search, product pages | 15 to 20% | Filters, variants, catalogue size |
| Basket, checkout and payment | 15 to 20% | Payment methods, delivery, click & collect |
| Account, loyalty, history | 10 to 15% | Points programme, in-store card |
| Notifications and segmentation | 5 to 10% | Simple sending or automated scenarios |
| Connection to the existing back office | 10 to 20% | Quality of your platform's API |
| Testing, publication, going live | 10 to 15% | Number of devices, store requirements |
The typical scopes we offer, all on a free quote within 24 hours and at low prices:
- MVP on a single platform: the most economical option, in 4 to 8 weeks. Catalogue, basket, payment, account, transactional notifications.
- Full iOS + Android app: in 3 to 4 months. Loyalty, segmentation, favourites and alerts, in-store features.
- Maintenance: a monthly fee, depending on the pace of change. It is not optional: without updates, an app becomes unstable and then impossible to publish with new versions of iOS and Android.
The fixed store costs remain modest: an annual Apple developer account subscription and a one-off fee for Google Play, a few tens of euros in total. And good news that is often overlooked: Apple and Google take no commission on the sale of physical goods. You keep your usual payment provider. For the detail of our plans, see our pricing or our article on the cost of a mobile app in 2026.
The five mistakes that make an e-commerce app fail
1. Copying the site identically. If the app does exactly the same thing as the site, nobody has a reason to install it. It must bring at least one concrete advantage: faster payment, a loyalty benefit, a stock alert, early access.
2. Not promoting it. An app does not get discovered on its own in the stores. A smart banner on the mobile site, a mention in confirmation emails, a QR code in store and on parcels: this is the work that fills the user base.
3. Letting stock and prices get out of sync. A product available in the app but actually sold out destroys trust faster than any bug. The app must read the same source of truth as the site, in real time.
4. Forgetting measurement. Without tracking purchases by channel, it is impossible to know whether the app pays off. Define from the start the three indicators that matter: install rate among customers, purchase frequency of app users, additional revenue.
5. Building everything in version 1. Start with the perfect purchase journey and transactional notifications. Advanced loyalty, augmented reality or live shopping can wait until the data proves they have an audience. Our method for scoping an MVP applies here without modification.
How to decide, in summary
- Your mobile site converts poorly: fix it first, it is the best possible return on investment.
- You have few repeat customers or a rare purchase: an excellent mobile site is enough, possibly enriched as a PWA.
- You want to test the appetite for an installed channel on a limited budget: a PWA or an MVP on one platform.
- You have a large base of customers who buy several times a year, or physical stores: the native app becomes a profitable investment, not a showcase.
At 55 · agency, we have been building apps and websites since 2013, for retailers such as Decathlon Camp and brands such as Yellow Kitchens. On e-commerce projects, our first job is often to say whether the app is justified, before talking about features. You will find our detailed approach on the e-commerce mobile app page (in French), and the main stages of a project on mobile app development.
Hesitating between mobile site, PWA and native app? Send us your key figures (repeat customers, purchase frequency, current platform) via our contact page or on +33 1 59 13 25 55: we will tell you in one conversation which option makes sense for your shop, even if the answer is "not yet".
Sources
- Progressive Web Apps — Google (web.dev)
- App Review Guidelines — Apple
- Become a member — Apple
- Premiers pas avec la Play Console — Google Play Console
