How to Build an E-Commerce Platform Like Shopify: Storefront Engine, App Ecosystem, and What Merchants Need

Most founders who come to us wanting to "build something like Shopify" are not trying to compete with a $100 billion company. They are building a vertical commerce platform for a specific industry that Shopify serves badly: regulated goods, event ticketing, B2B wholesale, food service, rental platforms, or a geography where Shopify's payment coverage is thin. The platform shape is similar. The go-to-market is completely different.

A multi-tenant e-commerce SaaS MVP costs $80,000 to $160,000 and takes 20 to 30 weeks. Here is what drives the range:

The drag-and-drop builder is the single biggest scope variable. It is a visual editor. That is a separate product in itself.

According to Shopify's 2023 Annual Report, the platform processed $235 billion in GMV that year. But Shopify serves primarily the small-to-mid market. The verticals with compliance requirements, complex pricing models, or non-standard business logic are where custom platforms win.

How Shopify makes money (and what your model should be)

Shopify's revenue in 2023 was split roughly 70% merchant solutions (payment fees, fulfillment, capital) and 30% subscription revenue (monthly plans from $39 to $399). The platform earns more from processing transactions than from plan fees. That is the model to study if you are building a vertical equivalent.

For a vertical commerce platform, the standard monetization structure is:

Monthly subscription per merchant. Most vertical platforms charge $99 to $499 per month depending on tier and feature access. This is your predictable base revenue.

Transaction fee on GMV. Typically 0.5% to 2% of every transaction processed through your platform. At $10 million in annual GMV across merchants, a 1% fee generates $100,000 in platform revenue. This is where margin lives as you scale.

Payment processing spread. If you own the payment layer (Stripe Connect), you earn the difference between what Stripe charges you and what you pass to merchants. Standard spread is 0.3% to 0.5% on card volume.

Optional: setup fees, premium themes, priority support. These are secondary but worth including in your model.

Unit economics at 100 merchants averaging $5,000 monthly GMV each: $500,000 in GMV per month. At a 1% platform fee plus $199 average subscription, you earn $5,000 in transaction fees plus $19,900 in subscriptions per month. That is roughly $300,000 in annual platform revenue before infrastructure costs.

At 500 merchants with the same GMV average, annual platform revenue crosses $1.5 million. That is the threshold where the platform becomes interesting to investors and justifies deeper feature investment.

Who actually builds a platform like this instead of using Shopify

Four specific types of companies build vertical commerce platforms instead of directing merchants to Shopify:

Healthcare and regulated goods operators who run platforms for CBD, supplements, alcohol, pharmaceuticals, or adult products. Shopify can and does suspend merchant accounts in these categories. A vertical platform serving licensed dispensaries or supplement brands can build KYC into onboarding, handle age verification at checkout, and never face a policy-driven shutdown. The compliance cost is real but it is a moat, not a burden.

B2B wholesale platforms serving buyers who need net-30 credit terms, tiered pricing by volume, quote requests, and company-level accounts with multiple buyers. Shopify's model is built around consumer checkout. Wholesale buyers do not want a cart; they want a PO workflow. None of Shopify's plans handle this without significant workarounds.

Industry-specific platform businesses where the workflow is inseparable from the commerce layer. Event ticketing platforms (seat selection, PDF tickets, scanning at entry, transfer policies), food service platforms (time-slot ordering, prep routing to kitchens, driver dispatch), and equipment rental platforms (availability calendars, damage deposits, return scheduling) all need commerce infrastructure woven into a domain-specific workflow. Shopify apps can get close, but the seams show.

Geographic operators in markets where Shopify's payment support is limited or where local regulatory requirements demand a domestic payment flow. Southeast Asia, parts of the Middle East, and Latin American markets with specific local payment methods are common examples. Building a regional commerce platform with local payment provider integration gives merchants in those markets options Shopify cannot.

What the platform is actually made of

A Shopify-equivalent is not one app. It is four interconnected products running from a single codebase:

The merchant portal is where merchants manage their store: products, inventory, orders, customers, discounts, and settings. This is the primary surface your merchants use daily. It needs to be fast and obvious because most of your merchants are not technical.

The storefront layer is what shoppers see. For v1, this means 3 to 5 templated themes that merchants configure without code. Colors, fonts, logo, layout. Most merchants need configuration, not customization.

The commerce engine is the backend logic: inventory tracking, pricing rules, discount stacking, tax calculation, and checkout flow. This is where most of the edge cases hide.

Payment infrastructure connects customer payments to merchant payouts through your platform fee collection. This layer has more complexity than it looks.

How much does multi-tenancy cost you, and why does it matter?

Multi-tenancy is the decision that affects every other cost in the build. You need a single platform installation to serve hundreds or thousands of merchants, each with isolated data.

The simplest and most common approach is row-level isolation: all merchants in shared database tables, every row tagged with a merchant identifier, all queries filtered by that identifier. This is how the majority of vertical SaaS platforms are built. It is the right starting point and scales to tens of thousands of merchants without architectural changes.

The alternative, giving each merchant their own database, costs more to operate and adds complexity to deployments and backups. It makes sense for enterprise platforms where merchants have SOC2 or HIPAA requirements. If that describes your market, budget an extra $30,000 to $50,000 for the infrastructure layer and operational tooling.

The failure mode we see most often: teams start with row-level isolation, then discover mid-build that two enterprise prospects require database-level isolation, and spend six weeks retrofitting the data layer. If you have any chance of serving compliance-heavy merchants, make the isolation decision before writing the first query.

What features does a v1 platform need?

Most vertical commerce platforms scope too much for v1. The result is an 18-month build that validates nothing. Here is a disciplined breakdown:

V1: launch (20-30 weeks, $80K-$160K)

Merchant signup and onboarding: account creation, Stripe Connect setup, domain connection. Without this, merchants cannot start. The Stripe onboarding KYC flow alone takes 2 to 3 weeks to build and test properly.

Product catalog: products, variants, images, pricing, and basic inventory. Digital products need a separate download delivery path. Services need to suppress shipping.

Templated storefront: 3 to 5 configurable themes. Not drag-and-drop. Merchants configure; they do not design. This covers 80% of what merchants need at launch.

Checkout: shipping address, method selection with live carrier rates, payment, order confirmation email. According to Baymard Institute's 2024 cart abandonment research, the average documented cart abandonment rate is 70.19%. A checkout that adds friction on top of that number makes merchants leave. Build this carefully.

Order management: order list, order detail, fulfillment tracking, refund initiation. Merchants need to see and act on orders. That is the minimum.

Basic analytics: GMV per day, orders per day, top products. Not a full analytics suite. Just enough for a merchant to know if their store is working.

V2: growth (add after validating v1, roughly $55K-$90K additional)

Drag-and-drop storefront builder. This is the most requested feature and the most expensive to build. Budget $55K to $90K for a block-based visual editor. Launch with templates first, add the builder after you know what merchants actually want to customize.

Discount and promotion engine. Percentage discounts, fixed-amount discounts, buy-X-get-Y, free shipping thresholds. Merchants will ask for this within 90 days of launch.

Abandoned cart recovery. Automatic email sequence to recover abandoned checkouts. Direct revenue impact for merchants, so it builds retention. Add this after you have proven the core checkout flow.

Customer accounts and loyalty. Repeat purchase flows, order history, wishlist. Lower priority than getting first purchases working.

V3: scale (relevant above 500 merchants or $50M in platform GMV)

App marketplace. Open your API to third-party developers. This requires a stable, documented API and developer tooling. Do not open to external developers until the core API is stable enough that you are not making breaking changes.

POS integration for physical retail. If your vertical has brick-and-mortar merchants, this extends their online store into their physical location.

B2B wholesale module: net-30 terms, company accounts, quote workflows, volume pricing. Significant build, but it opens the door to a higher-value merchant tier.

International expansion: multi-currency, translated storefronts, local payment methods. Each market adds operational overhead. Expand to a second market only after the first is profitable.

How does payment collection work for the platform?

Stripe Connect is the standard architecture for marketplace and platform payments. The mechanics matter because they affect your cash flow timing and liability exposure.

Merchants onboard as Stripe connected accounts during signup. The KYC process (identity verification, business information) is handled by Stripe. Your job is to build a smooth onboarding UI around it.

When a customer purchases, the payment flows through your platform Stripe account. Your platform fee is deducted automatically. The remainder transfers to the merchant's connected account on your defined payout schedule: daily, weekly, or monthly.

Three things teams consistently underestimate here. First, the chargeback flow: Stripe handles disputes, but your platform needs to freeze merchant funds during a chargeback period and notify the merchant. Build this into the order state machine from day one. Second, refund policy decisions: when a customer returns something, does your platform fee get refunded too? That is a business decision with financial model implications, not a technical one. Make it before you build the refund flow. Third, payout reporting: merchants need to see exactly what they will be paid and when. A basic payout dashboard is not optional.

"Most of the interesting commerce infrastructure being built today is happening in the verticals Shopify ignores: regulated goods, complex B2B pricing, specific geographic markets, and niche business models. That is where platform builders find defensible ground."

-- Harley Finkelstein, President of Shopify (Shopify Unite, 2023)

Build vs. Shopify: when does custom win?

Keep directing merchants to Shopify when they are selling standard physical or digital products, their business model is straightforward, they are in a geography Shopify serves well, and they need the 10,000-app ecosystem Shopify has built over 15 years. Shopify's checkout conversion rates are difficult to match from scratch. If your merchants can use Shopify, they should.

Build your own platform when the following conditions are true:

Your vertical has compliance or data requirements Shopify cannot guarantee. Shopify's acceptable use policy explicitly restricts certain regulated product categories. A custom platform gives merchants in those categories a stable home without policy risk.

Your business model requires transaction economics Shopify does not support. Wholesale platforms need net-30 credit, quote workflows, and tiered pricing. Rental platforms need deposit holds, availability calendars, and damage fee collection. Auction platforms need bid management and reserve prices. None of these fit Shopify's checkout model without significant workarounds.

Your merchants need a workflow that is inseparable from the commerce layer. Ticketing platforms need seat selection wired into inventory. Food platforms need order routing to kitchen displays wired into checkout. These are not Shopify app problems. They are platform design problems.

Your geography requires payment or language support Shopify cannot provide. Building for a specific regional market with local payment methods, local tax requirements, and local-language merchant support is a legitimate reason to build.

According to a 2023 McKinsey study on platform economics, vertical SaaS platforms that target a specific industry with purpose-built workflows generate 2 to 3 times the customer lifetime value of horizontal platforms trying to serve all markets. Narrow is a strategy, not a limitation.

What RaftLabs builds in this space

We build vertical commerce platforms and multi-tenant SaaS products. The engagements we do most often in this category are: B2B wholesale platforms with net-30 credit and quote workflows, compliance-specific commerce for regulated categories, and regional platforms in markets Shopify serves poorly.

The first conversation is a scoping call: we review your merchant type, your business model, and your v1 feature set, then give you a realistic cost and timeline estimate within 48 hours. If you are building a vertical commerce platform, book a 30-minute scoping call.