— 14 July 2026
Does your business need an app, or just a website?
A practical decision framework for founders staring down a ₹5-30L build. Six questions that reveal what you actually need — before you spend a rupee.
Every second small business owner in India asks the same question. “Should I build an app?” The answer isn’t yes or no. It’s a set of six questions that reveal what you actually need — and what you can defer until you have revenue to justify it.
Question 1: Are users returning weekly, or just visiting once?
Websites win for one-time visitors — service pages, marketing brochures, appointment bookings. Apps win when users come back three or more times a week — banking, food delivery, learning platforms, dealer portals.
If your buyers touch you monthly or quarterly, a website plus WhatsApp is almost always enough.
Question 2: Are users on mobile, or on a laptop?
Sounds obvious but plays into cost. If 90% of your users are on Android phones with 3G connections and mid-range hardware, a Progressive Web App or a Flutter mobile app beats a desktop-first website every time. If 60% are on office laptops (B2B admin tools, dashboards, back-office), start with web.
Question 3: Do you need to work offline?
Field sales, delivery routes, inventory in shops with patchy wifi, medical rounds — these need offline-first. That’s a mobile app. If your users are always online, save yourself the ₹8-15L of native development and ship a fast website.
Question 4: Do you need device features?
Camera, GPS, push notifications, biometric login, contact sync — these push you toward a mobile app. A PWA can handle some of it (push, camera basic), but if you need deep integrations, native or Flutter is the honest answer.
Question 5: What’s your discovery channel?
If you’re getting traffic from Google, Instagram ads, or referrals — a website is where they land. Apps require an install commitment your visitor may not make on first contact.
The pattern that works well for D2C brands: a fast marketing site for discovery, an app for retention. The site converts strangers into shoppers. The app converts shoppers into weekly repeat buyers.
Question 6: What’s your team going to update every week?
If your team wants to publish new products, blog posts, offers, events — that’s a website’s home turf. If they need to update dynamic pricing, live inventory, or interactive tools — that’s app territory (or a hybrid).
The typical answer
For 8 in 10 businesses under ₹5Cr revenue, the honest answer is: build a fast, well-designed marketing site with a strong contact/lead flow. Skip the app for now. Revisit in 12 months when you have paying customers who ask for one.
For the remaining 2 in 10 — mostly D2C brands with weekly repeat behavior, and B2B platforms with field operations — build the app. But build the website first anyway. Discovery still starts with search.
What we usually recommend
- Marketing site under ₹3L: WordPress or a static site
- Marketing site with a product configurator or booking flow, ₹3-8L: WordPress + custom modules, or Next.js
- Web app (dashboards, admin panels, ordering portals), ₹8-25L: Next.js + Supabase or Neon
- Mobile app (Android + iOS), ₹15-40L: Flutter + Supabase or Firebase
- Superapp (multi-module mobile + admin web), ₹30L+: Flutter + Next.js admin
Numbers move based on complexity. But this is the honest ballpark most Bengaluru studios quote in 2026.
When in doubt, ship the smaller thing first
The biggest mistake first-time buyers make is building the ambitious version before validating the small one. Ship the website. Watch what users actually do. Build the app in six months when demand is real, not hypothetical.
3 RELATED READING
Next to this one.
What we charge, and why
Real numbers for web and mobile builds in India — sprint, full build and retainer. What sits inside each band, what pushes a quote up, and how to compare three proposals fairly.
Read ↗What AI actually costs inside your product
Real token maths in rupees. What an AI feature costs per user per month, the four levers that cut the bill by 80%, and the guardrails that stop a runaway invoice.
Read ↗5 things that go wrong when you build software without a spec
The pattern shows up every kickoff week for a client who thinks they can 'figure it out as we go'. Here's what breaks — and the one-page fix.
Read ↗→ NEXT POST