Mobile App vs Mobile Site: Which Should You Build?

In Brief

The answer depends on your revenue, your repeat purchase frequency and who your customers are. Below roughly 500,000 TL in annual revenue, a fast mobile-friendly site is enough on its own. Once you pass 10,000 monthly active users and at least 30 percent of your customers place a second order, an app investment starts paying for itself. Between those two points, test demand with a PWA before committing money to a native build.

A pattern shows up constantly in fashion brand dashboards: 78 percent of visitors arrive on a phone, yet only 54 percent of revenue comes from mobile. The first idea raised in the meeting is usually an app. The same dashboard, though, also shows a checkout spread across four screens and a shipping fee that appears as a surprise on the final step. That is where you look before you change channels.

Venn Apps data puts 59 percent of worldwide e-commerce sales on mobile devices, with that share expected to reach 63 percent by 2028. Figures compiled by Capital One Shopping show 85 percent of US mobile shoppers preferring apps to mobile sites. Turkey is sharper still. Years of marketplace shopping mean the average buyer learned how to order inside an app long before your brand asked them to install one.

Timing is where money gets lost. Setting aside a 300,000 TL app budget before you have steady traffic is like dressing the window of an empty store. The reverse mistake costs just as much: a mature brand with high repeat purchase frequency and no app hands its most valuable customers to a competitor's notification feed. What follows compares the two options on cost, conversion, loyalty and daily operations.

The Mobile-Friendly Site: An Open Door

A responsive site is a single codebase that adapts automatically to phone, tablet and desktop screens. Nothing gets installed, no account is created, no permission is requested. The visitor taps and is inside.

Reach and Discoverability

Frictionless access is the responsive site's real strength. Someone searches for a black trench coat, taps a result and lands on your product page within three seconds. Almost all new customer acquisition happens through that path. There is no alternative on the SEO side either, because search engines index your web pages rather than the screens inside your app. Data published by JMango360 shows mobile websites beating apps on new customer acquisition cost precisely because they are easier to discover. The same logic extends to AI-assisted search: if your catalogue is not crawlable, those systems cannot surface it.

Cost and Update Speed

You build one platform and run it everywhere. In Turkey, responsive development for an e-commerce store typically costs 50,000 to 150,000 TL and takes two to four months. Changing a campaign price never waits for store approval; the edit goes live immediately. On a well-chosen platform, mobile responsiveness arrives as standard rather than as a module you pay for later, which is why avoiding the costly platform selection mistakes shrinks your future mobile budget directly.

Where It Falls Short

• Conventional push notifications are unavailable, closing off the cheapest way to bring a customer back.
• It does not work offline. Lose the signal on the metro and both the cart and the browsing history disappear.
• There is no full access to the camera, location services or motion sensors.
• The experience is less fluid. MobiLoud's analysis found shoppers spending an average of 201.8 minutes a month in apps against 10.9 minutes on mobile sites.

On its own, a responsive site is sufficient for early-stage businesses, small companies with tight budgets and categories with low repeat purchase frequency. A customer buying furniture, appliances or building materials orders once a year. Expecting that person to install an app is not realistic.

The Mobile App: Loyalty and Deep Engagement

A native app is software the customer installs from the App Store or Google Play. It reaches every capability of the phone and delivers the smoothest experience available.

The Conversion and Revenue Gap

Conversion is the app's clearest return. According to Capital One Shopping, mobile apps convert about three times better than mobile sites: 3.5 percent against 1.2 percent. Shoppers also view 4.2 times more products per session inside an app. JMango360 reports that app users spend roughly 30 percent more than web users, at 95 dollars per transaction against 73 dollars. None of that is accidental. The card is saved, the session is authenticated, the address is selected, and the customer opened your app deliberately. The methods for raising conversion rate work in both channels, but the ceiling sits far higher in an app.

Push Notifications and Retention

Push is the cheapest way to bring a customer back without spending on media. The coat left in the cart, the size 38 back in stock, a birthday coupon: each costs nothing per message. Fullstory's report found that 64 percent of consumers prefer a brand's app to its website, and that session lengths grew 332 percent in a single year. This channel poisons easily. A brand sending two notifications a day pushes its uninstall rate up sharply inside three weeks. Treat push as a personal reminder system rather than a campaign megaphone.

Device Capabilities and Cost

Camera, location, biometric login and background sync only work fully here. Scanning a QR code for in-store pricing, offering delivery from the nearest branch, letting someone try a product in augmented reality: those belong to the app. MobiLoud reports that retail apps convert 94 percent better than mobile sites. The invoice is heavy in return. A typical e-commerce app needs 200,000 to 500,000 TL and four to eight months of work. A single codebase built with React Native or Flutter cuts that figure by 30 to 40 percent, though never to zero.

PWA: The Step Before Going Native

A Progressive Web App is built with web technology but behaves like an installed app. The customer adds your site to the home screen, an icon appears, and the page opens full screen with no address bar. There is no store download step.

What It Solves and What It Does Not

Offline caching, home screen installation and Android push notifications all come with a PWA. Venn Apps expects PWA adoption to grow around 10 percent through 2026. The gaps are equally clear: no shelf space in the App Store or Google Play, restricted notification and background capability on iOS, and no social proof from store ratings. Those gaps get louder as your scale grows.

Who This Step Suits

• Brands in the 3,000 to 10,000 monthly active user range, where a native build is still premature.
• Teams that want to test app demand without risking 300,000 TL to find the answer.
• Merchants whose traffic skews heavily to Android, with a small iOS share.

Adding a PWA typically costs 20,000 to 50,000 TL and takes one to two months, roughly a tenth of a native build. The practical method: convert the site, then track three numbers for six months. Home screen install rate, notification opt-in rate, and the 90-day repeat purchase rate of the people who installed. If installers repurchase noticeably more often than everyone else, demand for a native app is proven. If they do not, your problem is the product or the price rather than the channel, and an app will only repeat that problem in a more expensive place.

Comparing Both Channels on the Same Criteria

Putting the two options on identical rows shortens the argument. The question is never which is better in the abstract, but which one wins on the metric you have chosen to prioritise.

Criterion	Mobile-Friendly Site	Mobile App
Ease of access	Very easy, no download	Download and install required
SEO strength	High, ranks in Google	None, requires ASO
Development cost	50,000 - 150,000 TL	200,000 - 500,000 TL
Development time	2 - 4 months	4 - 8 months
Update speed	Publishes instantly	Store approval required
Push notifications	None, limited via PWA	Full support
Offline use	None, limited via PWA	Full support
Conversion rate	Moderate, 1.2 - 2 percent	High, 3.5 percent and above
Revenue per transaction	About 73 dollars	About 95 dollars
Access to device features	Limited	Full access
Annual maintenance	Low	High, two platforms

Number of Users or Value per User

A site reaches more people; an app keeps the more valuable ones. Megh Technologies' analysis found mobile apps generating 80 percent of e-commerce revenue while web apps account for 20 percent. That average comes from brands that already have an app and a mature user base. For a brand starting from zero, the app stays a five percent channel for a long stretch, and the site carries everything during that period. Pick your metric according to your growth stage. During acquisition, traffic and cost per first purchase decide; during retention, repeat purchase rate and customer lifetime value decide. Setting up your ROAS calculation properly is the most practical way to separate what each channel really contributes.

Which Solution at Which Revenue Stage

Your revenue, your repeat purchase frequency and your active user count are the most reliable signals. The thresholds below work as a practical roadmap for typical e-commerce operations in the Turkish market.

Early: 0 to 500,000 TL a Year

Winning customers and validating product-market fit come first. A native app is premature. Put the budget into a fast responsive site, better product photography and advertising tests. Getting your pages to open in under 2.5 seconds on mobile is worth more at this stage than anything an app could add.

Growth: 500,000 to 5 Million TL a Year

When traffic and sales climb, loyalty enters the conversation. Move to a PWA first, then watch two thresholds: have you crossed 10,000 monthly active users, and have more than 30 percent of your customers placed a second order? Both true means the return on an app investment is positive. Only one true means collecting six more months of data.

Mature: 5 Million TL and Above

An app stops being a preference at this scale and becomes a competitive requirement, since most of your rivals are already on the stores. Loyalty programmes, member-only early access, personalised recommendations and augmented reality try-ons only earn their cost here.

Category Effects

• Fashion and cosmetics: high repeat purchase, strong visual experience. The app is decisive.
• Food and rapid delivery: several orders a week, notifications essential. The app is mandatory.
• Electronics and durables: infrequent orders, large baskets. The site suffices far longer.
• Furniture and building materials: one or two purchases a year. An app investment is unnecessary.

Megh Technologies also notes that the 25 to 34 group leads mobile shopping, and that 92 percent of people aged 30 to 49 shop from a smartphone. If your audience skews older, write their resistance to installing apps into the budget plan now.

Cost and the Payback Calculation

This is an arithmetic decision, not an emotional one. A five-row table ends an argument that has been running through meetings for months.

A Concrete Scenario

Picture a fashion brand with 40,000 monthly mobile visitors, an average basket of 1,200 TL and a mobile conversion rate of 1.4 percent. Assume 12 percent of those visitors migrate to the app, where conversion climbs to 3.2 percent.

Line item	If they stayed on the site	Group that moved to the app
Monthly users	4,800	4,800
Conversion rate	1.4 percent	3.2 percent
Monthly orders	67	154
Average basket	1,200 TL	1,200 TL
Additional monthly revenue	-	104,400 TL
Contribution at 35 percent gross margin	-	36,540 TL
Payback on a 350,000 TL build	-	about 10 months

Line Items That Never Make the Model

• Annual maintenance and version updates across two platforms: 15 to 25 percent of the initial build.
• Apple and Google developer accounts plus continuing store compliance work.
• Install acquisition cost, since driving people to the store needs its own media budget.
• Extra team hours to run content and campaigns separately in two channels.

Because of those items, an app gets built on top of your site rather than instead of it. Launching one without a solid e-commerce infrastructure underneath is like adding a floor to a building with weak foundations. When stock, pricing and campaign logic are not managed from one source, the inconsistency between channels costs you customers. Show a price 40 TL higher in the app than on the site and the person who noticed will not try a second time.

The Turkish Market and the Hybrid Setup

Several local realities change this decision in Turkey, and plans that ignore them do not survive contact with the market.

Marketplaces, Consent, Payments and Shipping

• Trendyol and Hepsiburada already satisfy the customer's app habit. Whatever your own app offers has to be something the marketplace cannot give: brand experience, member pricing, loyalty points that accumulate.
• In-app push falls outside the IYS definition of a commercial electronic message, since consent is captured at device level. SMS and email do require IYS registration. On the KVKK side, the privacy notice, explicit consent and retention periods must be defined clearly in both channels.
• Instalment tables for local cards, bank campaigns and BKM rules have to work in full inside the app. An app that hides instalment options converts worse than your website does.
• Tracking integration with local carriers becomes the most-opened screen in the app. If order tracking is not wired properly into logistics and warehouse management, the app turns into a complaints channel.

Building the Hybrid Model

Successful brands do not pick a side; they give each channel a different job. The site handles discovery and the first purchase, absorbing traffic from search, ads and social. The app handles retention and takes over after the second order. Engineer the handoff with incentives. Three placements consistently work: a post-purchase message offering 10 percent off in the app, a narrow banner on the home page, and a reminder just above the checkout step. Cart, wishlist, order history and account details have to stay synchronised across both, and deep links should drop the customer straight onto the relevant product. For businesses not yet at that scale, solutions built for smaller teams make it possible to establish the same foundation on one responsive platform. Here is the job for tomorrow morning: open your analytics and write two numbers on one line. The share of customers who placed two or more orders in the last 90 days, and your mobile page load time. Those two figures will settle where your mobile budget goes in a single meeting.

Frequently Asked Questions

Which is more profitable, a mobile app or a mobile site?

The site wins in the short term and the app wins in the long term. A site reaches a wide audience cheaply and brings in new customers. An app costs more but produces roughly three times the revenue per user. What actually decides profitability is how often your existing customers come back.

If I build an app, do I still need a website?

You certainly do. Search engines and AI search systems see your web pages, not the screens inside your app. The customer who eventually installs the app is almost always won by your site first. An app is a layer added on top, never a replacement.

Can a PWA really replace a native app?

Not entirely, though it covers a large share of real use cases. It gives you offline browsing, home screen installation and push notifications on Android. Its main gaps are the notification and background restrictions on iOS and the absence of app store visibility. As a way to test the investment, it remains the right first step.

How can I reduce app development costs?

A single codebase with React Native or Flutter cuts costs by 30 to 40 percent. Shipping a first version limited to purchasing, order tracking and notifications shrinks the budget further. Leave augmented reality, gamification and live support for the second release.

How many installs justify the investment?

Behaviour decides this, not the raw install count. Passing 10,000 monthly active users while more than 30 percent of your customers repurchase means you have crossed the threshold. Without both conditions, even a high install number will not produce a profitable app.

Do push notifications require IYS registration in Turkey?

In-app push notifications do not fall under the IYS definition of a commercial electronic message, because consent is granted at device level. SMS and email do require IYS registration. Your KVKK disclosure obligations continue to apply in both channels regardless.

I sell on Trendyol, do I still need my own app?

A marketplace binds the customer to the platform rather than to you, and neither the contact permission nor the behavioural data stays with your brand. Your own app buys independence in loyalty and repeat sales. If marketplace volume has not yet turned into regular repeat purchases, strengthen your own site first.

How long does it take to publish an app?

Development runs four to eight months, and store review usually takes a few days to two weeks. Rejected builds add correction cycles on top of that. For that reason, avoid tying a launch tightly to a fixed campaign date.

Will an app fix my slow mobile site?

No, in most cases it inherits the problem. Slowness usually starts in the server, the image handling or the data model, and the app calls exactly the same services. Fix site speed first, and hold the app budget until you can show measurable improvement.

I do not have a team to run both channels, what should I do?

In a one-person operation, running two channels separately produces content inconsistency within weeks. Do not move to an app before you have infrastructure feeding product, price and campaign data from a single panel. Until that setup exists, staying on the responsive site is cheaper and less risky.