Every Sharetribe project I've seen eventually reaches the same meeting. Someone on the founding team says, "We need an app." A developer asks, "Which problem does the app solve that the responsive site doesn't?" And the room goes quiet.
This post is for the developer in that meeting. It treats "web, mobile or both" as an engineering decision with inputs, signals and a rollout plan, not as a branding question. The ideas come from the guidance that Journeyhorizon published on whether a Sharetribe marketplace needs both a mobile app and a web marketplace, reworked here into something you can bring to a sprint planning session.
Start with what each surface is actually for
Before comparing costs, get the job descriptions right. The two surfaces sit at different points in the funnel.
The web marketplace: discovery, trust and validation
A Sharetribe web marketplace is the browser version of the product, reachable by URL from any device. It's usually the first thing you ship, and for good reasons:
- It's the fastest way to validate supply, demand and positioning.
- It's the easiest surface to edit and redeploy.
- It's the only surface search engines can crawl.
That last point matters more than most teams admit. Niche marketplaces get found through Google, social posts, referrals, ads, communities and content. An app store listing doesn't replace a category page for "equipment rental marketplace" or a city-specific landing page for a service. Rental platforms with per-city category pages, service marketplaces with pages per service and location, and B2B marketplaces with case studies and educational content all depend on crawlable HTML.
The mobile app: speed, frequency and retention
A Sharetribe mobile app is an iOS and Android client talking to the same marketplace backend. It covers browsing, search, messaging, booking or purchasing, profile management and notifications. Its advantages are mobile-native: push notifications, camera access, location and persistent sessions.
It also brings a new operational surface: store accounts, build pipelines, release management, device QA, Apple and Google policy compliance, and maintenance that never ends.
A useful mental model: web is top of funnel, mobile is middle and bottom of funnel. Apps can still acquire users through ASO, paid campaigns, referrals and brand search, but for most niche marketplaces the website stays the main acquisition channel early on.
The comparison, as a config table
Here's the trade-off space collapsed into something you can paste into a design doc:
web_only:
best_for: validation, SEO
growth_role: acquisition + trust
seo_value: strong
cost: usually the cheapest start
launch_speed: fastest
choose_when: still validating supply, demand, positioning
custom_mobile_app:
best_for: repeat users, push alerts
growth_role: retention
seo_value: limited
cost: can get expensive
launch_speed: slower
choose_when: users need frequent access, alerts, messaging, mobile booking
web_plus_mobile_template:
best_for: growing marketplaces needing SEO + retention
growth_role: full funnel
seo_value: strongest (SEO stays on web)
cost: lower upfront, custom work added selectively
launch_speed: faster than a custom build
choose_when: traction exists, mobile demand is clear
Don't ship both on day one (usually)
Launching both surfaces at the start splits your attention and spends budget before you have usage data. A web-first launch lets you test categories, see which listings convert, refine onboarding and sharpen positioning, all on the surface that's cheapest to change.
There are real exceptions. Some marketplace models are built on high-frequency mobile behaviour:
- local services
- rentals
- delivery
- event-based platforms
- fitness or activity booking
- peer-to-peer platforms
If your core loop depends on instant messaging, location-based browsing, push reminders, fast rebooking or uploading photos from the field, mobile is closer to the centre of the product. The rule is to let user behaviour make the call, not a competitor's app store badge.
Debug the mobile web before you build an app
Before anyone opens Xcode, run this checklist against your responsive site. A lot of "we need an app" requests are really "our mobile web flow is broken" bugs.
- [ ] Is there sign-up drop-off on mobile specifically?
- [ ] Is messaging on a phone awkward enough that replies get missed?
- [ ] Does booking or checkout take too many taps on a small screen?
- [ ] Can providers manage listings, availability and orders from a phone?
If you can fix these in the web client, you may have bought yourself months before an app becomes necessary. Web alone is often enough when the marketplace is early, transactions are infrequent, acquisition is the bottleneck, or the model is research-heavy: B2B, high-value services and professional networks.
Five signals that justify a mobile client
When the data starts showing these patterns, an app moves from nice-to-have to next investment:
- Repeat usage. Users come back weekly, daily or several times a month.
- Messaging intensity. Slow replies cause missed messages or abandoned transactions.
- Provider operations. Sellers manage listings, bookings or availability on the move.
- Trust and legitimacy. Investors, enterprise partners or customers expect an app.
- Paid acquisition. A better mobile experience improves conversion and retention of users you paid to acquire.
Why a from-scratch app is a scope trap
A custom marketplace app isn't a set of screens. The real backlog looks more like this:
UX strategy, UI design, mobile frontend, API integration,
authentication, listings, search, messaging, payments,
booking, account settings, QA, deployment, analytics,
store compliance, ongoing updates
Marketplaces multiply this complexity with multiple user roles, transaction states, availability rules and edge cases. Custom development isn't bad, and as a marketplace grows it's often necessary. The failure mode is building everything before you know which features matter, which leads to wasted spend, launches delayed by months and a heavy maintenance load.
The template-plus-custom pattern
The better engineering pattern is to treat core flows as commodity and spend custom effort only on differentiation.
A Sharetribe mobile app template covers the baseline: accounts, listings, search, messaging, transactions, bookings or purchases, profiles and store deployment. None of those make your marketplace special; users simply expect them to work.
Your custom budget then goes to the features that do differentiate you:
- a unique onboarding flow
- advanced filters
- a custom booking calendar
- recommendations
- vendor dashboards
- special checkout rules
- integrations with business systems
This works because Sharetribe already provides a strong backend. The app should be a well-built client of that backend, not a reimplementation of it.
When to skip the template
Go fully custom when the mobile experience is the product: complex real-time workflows, deep third-party integrations, advanced maps, custom matching algorithms, offline functionality, complex multi-role dashboards, or a mobile-first core model.
Keep SEO on the web, permanently
Even after you ship an app, the website keeps its job. Crawlers need indexable pages; users and partners want a web presence before they install anything. Category pages, comparison pages, content marketing and local landing pages all belong on the web. Think of it as a permanent split of responsibilities: web drives discovery, mobile drives repeat engagement.
A rollout plan you can put in the roadmap
- Phase 1, web first. Validate, build supply, invest in SEO.
- Phase 2, mobile template. Add iOS and Android once repeat usage and mobile demand are visible in the data.
- Phase 3, targeted custom work. Build only the high-impact features the template doesn't cover.
- Ongoing. Keep evolving both channels.
And a final decision gate. Answer honestly:
- [ ] Do most users access the marketplace from a phone?
- [ ] Do they return often enough to justify an install?
- [ ] Are transactions lost to slow mobile messaging, booking or checkout?
- [ ] Do providers need to manage listings or orders from a phone?
- [ ] Would push notifications change behaviour?
- [ ] Is the category app-driven?
- [ ] Do users expect iOS and Android?
- [ ] Is there budget to maintain an app after launch?
Mostly "no" means keep improving the web client. Several "yes" answers mean an app is the logical next step, and if you're unsure, a template is the lower-risk way to find out.
Takeaway
"Do we need an app?" is a question with measurable inputs. Ship web first, watch for repeat usage and mobile friction, start mobile from a template, and save custom engineering for the features that set you apart. If you're planning that path on Sharetribe, the team behind this framework offers Sharetribe development and mobile app work and has been building marketplaces on the platform for years.
Top comments (0)