DEV Community

Harjot Rana
Harjot Rana

Posted on Fully Autonomous

What actually decides the cost of a marketplace app

A marketplace looks like one app, but it is really two. Buyers need one set of screens and sellers need another, and money has to move between them. That is why it costs more than a plain web app, even when the first demo looks simple.

Here is how I break it down when I scope one.

The two versions

1. A narrowed first version. Buyer and seller roles, one category, and one payment integration. The written scope confirms it fits one sprint.

2. A full two-sided build. Self-serve seller onboarding, payouts, disputes, and messaging.

In my scope estimator, the first one reads as $6,500 to $9,000 over about 4 to 5 weeks. The second gets no number until a written scope splits it into sprints. Those are estimates from fixed rules, not quotes. They assume a web app (not native mobile), three roles (buyer, seller, and admin), two or three workflows like listing and buying, one payment integration, standard customer data, and a public launch.

Why a full marketplace is scoped first

My Launch Sprint covers up to 3 core flows, sign-in, a database, and one payment integration or outside service. A full marketplace adds a second user type with its own onboarding, payouts, and trust features. That does not fit inside one sprint, so my sprint terms list two-sided marketplaces as out of scope. They get a written scope before they get a price.

What the narrowed version skips

Usually:

  • self-serve seller onboarding
  • automatic payouts
  • reviews
  • messaging
  • disputes

You or an admin handle those by hand while you learn whether buyers and sellers show up at all. The written scope lists what is in and what is out, so nobody is surprised later.

What pushes the cost up

  • Automatic seller payouts beyond basic checkout
  • Four or more workflows, like reviews or disputes
  • Identity checks or regulated payment steps
  • Live updates, queues, and retries at launch
  • A native mobile app on top of the web app
  • Several categories with different listing rules

What you can cut for a first version

  • One category and one city or niche
  • Onboarding the first sellers yourself
  • One checkout, sellers paid outside the app
  • A private beta with a small group first
  • Reviews, messaging, and disputes in sprint two
  • An admin who runs early orders by hand

Payment fees are a separate bill

Payment providers charge their own fees, billed to your account. A build price covers building and launching the app, not the fees on each order afterwards.

Check this before paying anyone

Hosted marketplace platforms are worth checking first if your model fits their template. If it does, you may not need a custom build yet.

A cheaper way to start

If the full build is more than you want to spend up front, a paid scope is the smaller first step: a one-page scope that names the exclusions, a fixed price, and a first prototype of the core flow. With me that is $500, credited to the build if you go ahead within 30 days.


The full breakdown for marketplace apps, with FAQs, is here: harjotrana.com/cost/marketplace-app. If you want a number for your own idea, the free scope estimator uses the same rules.

Happy to answer questions in the comments.

Top comments (1)

Collapse
 
wbcomdesigns profile image
Wbcom Designs 💡 •

"A marketplace is really two apps" is the right way to say it, and splitting v1 from the full build is honest scoping. The piece that blows estimates up in our experience is refunds and disputes: if sellers are paid instantly and a buyer disputes later, who eats it? Deciding that before building payouts changes the architecture.

For services marketplaces that start narrow, there are also off-the-shelf routes (Sharetribe on SaaS, or plugins such as WP Sell Services on WordPress) that can cover v1 before a custom build. Disclosure: we're Wbcom Designs and make WP Sell Services.