Solutions · Commerce & D2C

Marketplace app: what to decide, build and test.

Planning one? Here's what to settle before you brief anyone, what goes in the first version, how the build runs week by week, how the beta is tested with real people, and the numbers that show it's working.

Who touches the money, and other forks that set the whole design

A marketplace is really a set of decisions about trust and cash. Make these before anyone draws a screen.

  • Classifieds or managed checkout. In classifieds, buyer and seller meet and pay each other offline; you carry no payment risk but also no leverage on fraud. In managed checkout you collect, hold and pay out, which gives you trust tools and a revenue line, and also refunds, disputes and tax duties.
  • Commission, listing fee or subscription. Commission pays when sellers win; it needs managed checkout, or sellers will take deals off-platform. Listing fees and subscriptions work in classifieds, but sellers pay before they know it works.
  • Which side to win first. Pick the harder side and build for it. For a craft market like Gildscape that is artists; for resale it is buyers with real intent.
  • Curated or open. Reviewing every seller keeps quality up and growth slow. Open listing grows fast and needs moderation from the first day.

Checklist before the brief:

  1. Does money pass through you? Yes or no.
  2. How do you earn, in one sentence?
  3. Which side do you recruit first, and how many do you need to seed?
  4. Who reviews reports and listings every day?
  5. What is the one city or category to start in?

Version 1 scope: trust tools before growth tools

ModuleFirst versionAdd laterWhy
ListingPhotos, title, price, locality, category fieldsBulk upload, price hintsPosting speed decides supply.
SearchCategory, locality, price filtersSaved searches, alertsAlerts need a large catalogue to be useful.
IdentityPhone OTP for both sidesDocument KYC for high-value sellersOTP stops casual abuse; KYC adds friction.
ChatIn-app chat with scam warningsVoice calls, masked numbersKeeping talk in the app is your evidence trail.
PaymentsSkip in classifieds; held payment in managed checkoutSplit payouts, instalmentsPayment is the largest legal and support load.
PayoutsWeekly statement the seller can readDaily or instant payoutClear statements prevent most seller disputes.
ModerationReport button, queue, remove and restrictAuto-flags, image matchingHumans first, then automate what repeats.
ReviewsPost-deal rating, one per dealSeller badges and tiersBadges mean little before data exists.
InspectionOnly for high-value goodsPhoto checklists, hubsIn Secondgear inspection is the product, not an add-on.
AdminUsers, listings, reports, feesRevenue dashboardsYou need to act on abuse before you analyse it.

Eight to twelve weeks, and what you provide at each stage

  1. Weeks 1 to 2: rules and flows. We write the listing, report and payout rules with you. You supply a category list with the fields each needs, the fee model, and examples of scams you have seen.
  2. Weeks 3 to 4: design. Buyer, seller and moderator screens. You supply sign-off on the rules and a decision on language, such as Hindi alongside English.
  3. Weeks 5 to 8: build. Posting, search, chat, accounts, admin. You supply 30 to 50 seed listings with real photos, and test phones across price points.
  4. Weeks 9 to 10: payments and payouts, if managed checkout. You supply a payment aggregator account in your business name, bank details, and your CA's view on invoices.
  5. Weeks 11 to 12: pilot. Invite a small group in one locality. You supply named moderators who will clear the queue daily, and a support number.

A classifieds product can finish nearer eight weeks; managed checkout with payouts tends toward twelve. See the MVP guide for testing the idea first.

How we work with you each week

The same rhythm runs from week 1 to launch, so you always know what's happening and what we need from you.

  • Friday: a demo link and a 30-minute call. You open the latest build on your own phone and try it yourself, rather than watching our screen.
  • Within 48 hours: your feedback, in writing. In one shared doc, next to the screen it's about. Written feedback is clear, dated and doesn't get lost between calls.
  • Monday: triage. Every comment goes into fix now, next week or later. Anything that changes the agreed scope gets a written quote before we touch it.
  • Each milestone: two revision rounds, then your sign-off. Sign-off is against the scope we agreed, so it's clear what done means.
  • One decision-maker on your side. Most delays aren't build time. They're feedback waiting on several people who disagree.

The full version, including how the beta and user testing work across every kind of project, is on our process page.

The beta: a pilot in one locality

The marketplace beta is the pilot in weeks 11 and 12: one locality, one or two categories, real people on both sides. An empty marketplace teaches nothing, so supply goes in first.

  • What's in it. Posting, search and filters, in-app chat, the report queue and moderation tools, and held payments and payouts if you chose managed checkout.
  • What stays out. Saved-search alerts, seller badges, bulk upload, promoted listings. They need data and volume the pilot won't have.
  • Supply first. Before any buyer is invited, 10–15 sellers you recruited post 30–50 real listings with real photos. If you can't recruit 15 sellers by hand, that is the finding.
  • Then buyers. 15–20 buyers from the same locality who actually want that category, recruited by you, plus one or two named moderators who clear the queue every day.
  • How they get it. TestFlight or Google Play testing for the apps, a staging link for the admin and moderation console. A feedback button on every screen sends a screenshot and a note to the shared doc.

A new personal Google Play developer account has to run a closed test with at least 12 testers for 14 days before it can publish. Count the pilot group towards that, or use an organisation account.

What we test with real people, and the tasks we set

Each person gets a task on their own phone and we watch without helping. Five sessions per role usually surface the big problems. Both sides get tested, and so does the moderator: a marketplace fails as often in the admin as in the app.

WhoThe taskWhat we watch for
Seller"List this item with three photos and a price."How long does it take? Which category fields confuse them? Do photos upload on weak data?
Buyer"Find one under ₹2,000 near you and ask the seller a question."Do filters match how they think? What makes them trust a listing, or not? Do they stay in the app to chat?
Buyer, managed checkout"Pay for it and tell me when you'll get your money back if it doesn't arrive."Can they explain held payment back to us in their own words?
Seller"Read your weekly statement. How much did you earn after fees?"Can they work out the number? Every confused fee line becomes a support call later.
Moderator"Clear these ten reports. Two of them are scams."Speed and accuracy, and whether the tools let them remove, restrict and warn without asking us.

The numbers that say it's working

Pick two or three in week 1. We add the tracking before the pilot and report at launch and 30 days later. We don't promise the numbers; we make them visible so you can decide what to fix.

GoalWhat we measureWhere it comes from
Sellers list easilyTime to post a first listing; share of sellers who post a secondApp analytics
Buyers find thingsSearches with no results; listing views that lead to a chatApp analytics
Listings get demandShare of listings with at least one enquiry within 7 daysListings and chat data
Deals happenChats that turn into a deal or a paid orderChat and order data
Trust holdsReports per 100 listings; hours to clear the report queueModeration console
Both sides come backSellers and buyers still active after 30 daysAccount data

What a marketplace costs and what moves it

Our app cost guide gives small-studio ballparks. A classifieds product with buyer, seller and admin screens but no payments sits in the two-user-types-plus-admin band, ₹10–25 lakh. A marketplace with live maps and payments, such as one that includes delivery, sits at ₹25–60 lakh. These are ballparks, not a quote; we quote after a brief.

  • Money handling. Holding payments, splitting payouts and handling refunds is the biggest single step up.
  • Moderation depth. A report queue is modest; image matching, auto-flags and fraud scoring add real work.
  • Seller onboarding. OTP only is light; document checks, bank verification and approvals add screens and support.
  • Categories and attributes. Ten categories with different fields cost more than two.

Rules and habits for marketplaces in India

  • Collecting money for others. RBI issued its Regulation of Payment Aggregators Directions, 2025 on 15 September 2025. They state that a payment aggregator shall not carry out marketplace business, and that escrow accounts are for authorised aggregator business only. In practice: use a licensed aggregator for settlement rather than receiving seller money into your own account, and have a lawyer confirm your structure.
  • Tax collected at source. GST law lets the government set a collection rate of up to 1% for e-commerce operators on sellers' net taxable supplies; the rate was lowered to 0.5% from 10 July 2024 (0.25% CGST and 0.25% SGST, or 0.5% IGST). Whether it applies to you depends on your model, so your CA should confirm. Build the reporting in from the start.
  • Seller registration. Some small sellers making sales of goods within their own state through an operator have been exempt from mandatory GST registration since October 2023. Seller onboarding should record a GSTIN where there is one and handle the rest by policy.
  • Consumer rules. The e-commerce rules, amended in 2026 and in force from 1 January 2027, require marketplace entities to show key seller and product details and not to use consumer information without express consent. Plan fields for them.
  • Data. The DPDP Rules, 2025 follow an 18-month phased timeline and require verifiable consent for children's data. Decide whether under-18 users are allowed at all. See the DPDP checklist.
  • Language and habits. Sellers often post from a phone in Hindi or a regional language; see regional-language apps.

This is general information, not legal or tax advice. A marketplace touches payments, tax and consumer law, so take advice before launch.

Marketplace mistakes and what to do instead

  • Building both sides at once. Seed one side first, because nobody posts into an empty shop. Recruit 30 to 50 sellers by hand.
  • Leaving phone numbers in the page. Numbers in page source invite scraping. Reveal after OTP with a daily limit.
  • No moderator named. Reports pile up and trust collapses. Give one person a daily slot.
  • Taking payments in week one. Start with classifieds if you can; add managed checkout once deals happen.
  • Commission with no checkout. Sellers move the deal to WhatsApp and you earn nothing. Match the fee to what you control.
  • Too many categories. Choose one category in one locality, then widen.
  • Skipping the website. Buyers arrive from search and shared links, so launch on the web first. See the marketplace guide and the second-hand goods playbook.

Cost and timeline

8–12 weeks for a first version.

Fixed quote within 48 hours of your brief: scope, timeline and price in writing before we start. These are what move the number.

  • Classifieds (buyers and sellers meet) or managed checkout (you take payment)
  • Seller onboarding, KYC and payout rules
  • Chat, calls and moderation depth
  • Inspection, delivery or escrow steps
  • Number of categories and attributes

How app costs work in India →

Questions

About your marketplace app.

How much does a marketplace app cost in India?

It depends on whether you only connect buyers and sellers or also handle payment, payouts and delivery, and on how much moderation you need. We give a fixed quote within 48 hours of your brief. See the app cost guide for the drivers.

How do you stop scams and fake listings?

We design trust in from the start. In our concept build Vendly, numbers stay hidden until a buyer verifies by OTP, chats flag common scam scripts, reused photos are caught, and reports land in a moderation queue. You can try each case on the Trust page.

Should we launch with an app or a website first?

Most marketplaces launch with a fast mobile website because buyers arrive from search and shared links, then add an app for repeat sellers. Our website or app guide covers the trade-off. A website also gets indexed by Google sooner. The app comes later.

How long does it take?

A first version with buyer, seller and moderation screens usually takes 8 to 12 weeks. A two-week concept sprint can test the idea with real users first; see our MVP guide. That way you learn which side is harder to win before spending on the full build.

Who owns the platform?

You do. The code, designs and data model are handed over in full, with documentation your team can work from. Listings, user accounts and transaction history stay in your own hosting, and any engineering team can extend the platform after hand-over.

Start a project

Tell us how your business runs. We'll build what it runs on.

Four taps, about a minute. Fixed quote within 48 hours.

Menu

Read

Ask an expert Ask an expertFree, real peopleHub
Ask an expert