Shopify vs WooCommerce vs Custom in 2026: A Decision Framework That Doesn't Lie

Every six weeks we get the same call: a founder or a head of digital, sometimes a CTO, asking whether they should launch on Shopify, rebuild on WooCommerce, or commission a custom stack. The honest answer is almost never the one they're hoping for, and it almost never matches what the last agency told them.

This is the framework we actually use when we sit down with that decision. No vendor loyalty, no "it depends" cop-outs — just the questions that matter and the cost lines that surprise people.

Start With Order Volume and Margin, Not Features

The single biggest mistake we see is teams comparing feature checklists before they look at unit economics. A platform decision is a multi-year cost commitment, and the math changes radically across three thresholds:

• Under ~2,000 orders/month, sub-$50 AOV: Platform fees and dev time dominate. You cannot afford custom.
• 2,000 – 50,000 orders/month: Shopify and WooCommerce are both viable. The decision turns on team skills and integration surface.
• Above 50,000 orders/month, or B2B with complex pricing: Shopify Plus or a custom/headless stack starts to pay back. WooCommerce gets painful here unless you have a serious ops team.

Layer margin on top. A 70% gross margin DTC brand can absorb Shopify's 0.15 – 0.6% transaction fee on Plus without flinching. A 12% margin electronics reseller cannot, and that single line item often forces them off Shopify regardless of how nice the admin is.

The hidden cost nobody quotes

When we model total cost of ownership over three years, the platform license is rarely the largest number. In our experience it tends to break down roughly like this:

• Platform/hosting fees: 10 – 25%
• Apps, plugins, and SaaS integrations: 20 – 35%
• Engineering and design (build + maintenance): 40 – 60%
• Payment processing: separate, but watch the gateway lock-in

WooCommerce wins on line one and loses on line three more often than people expect, because plugin conflicts and self-hosted infrastructure eat hours that Shopify quietly absorbs.

Shopify in 2026: What It's Actually Good At

Shopify has spent the last three years closing the gaps that used to push teams headless or custom. Checkout Extensibility is stable, Hydrogen and Oxygen have matured, B2B on Plus is finally credible, and Markets handles multi-currency without the contortions of 2022.

Where Shopify genuinely wins in 2026:

• Time to first revenue: A competent team ships a credible storefront in 4 – 8 weeks.
• Checkout conversion: Shop Pay's one-tap checkout consistently outperforms anything we've built custom, especially on mobile.
• PCI and fraud: Offloaded. This is worth more than people admit until they've handled a chargeback dispute themselves.
• App ecosystem: For 80% of the things a mid-market brand needs (reviews, loyalty, subscriptions, shipping), there's a $20 – $200/month app that works.

Where it still hurts:

• Liquid templating feels dated when your front-end team lives in React/Vue. Hydrogen helps but adds operational weight.
• Complex B2B pricing rules (customer-specific catalogues, tiered pricing with approval workflows) are doable but fragile.
• Transaction fees if you don't use Shopify Payments. In some regions Shopify Payments isn't available, which silently disqualifies the platform.
• Data ownership and export: getting your historical order data out in a usable shape is more work than the marketing suggests.

WooCommerce: Cheaper Sticker, Different Bill

WooCommerce is the option engineers either love or have been burned by. Both reactions are valid.

It makes sense when:

• You already run WordPress for content and the commerce surface is secondary (think publishers, course creators, B2B catalogs).
• You need deep customization of product data models without paying for Shopify Plus.
• You're in a market where Shopify Payments doesn't operate and Stripe/local gateways via Woo plugins are cleaner.
• Your team has real PHP and DevOps capability.

It becomes a liability when:

• You have no in-house ops to manage hosting, backups, plugin updates, and security patching.
• You're chasing >100 orders/hour during launches. Woo can do it, but tuning MySQL, object caching, and queueing background jobs is a specialist sport.
• Your plugin stack drifts past 25 – 30 active plugins. Conflict debugging becomes a part-time job.

The honest summary: WooCommerce's licence is free, but you're trading SaaS fees for engineering hours. If you don't have those hours, the bill arrives anyway, just labelled differently.

Custom and Headless: When the Math Finally Works

We build custom commerce stacks for maybe one in fifteen clients who ask about it. The other fourteen are better served by Shopify or Woo. Here's when custom is genuinely the right call:

1. Your commerce logic is the product. Marketplaces, rental, ticketing, configure-to-order industrial goods, multi-vendor B2B with complex approval chains.
2. Regulatory or data residency requirements force you off US-hosted SaaS.
3. Integration surface is enormous — ERPs, WMS, custom logistics, dealer networks — and the SaaS APIs become the bottleneck.
4. You are doing enough volume that 0.5% of GMV exceeds your engineering team's loaded cost. This is a real threshold and most brands never reach it.

Headless-on-Shopify (Hydrogen, or Next.js + Storefront API) sits in between. It's worth it when you have a content-heavy storefront, multiple brands sharing infrastructure, or a need to ship UI faster than Liquid allows. It's not worth it just because React feels nicer than Liquid.

A rough decision tree

If monthly orders < 2,000:
  -> Shopify (Basic/Advanced) or WooCommerce if WP already in stack

Else if standard DTC, AOV $20-$500, <5 ERP integrations:
  -> Shopify (Advanced -> Plus as you scale)

Else if complex B2B pricing OR heavy WordPress content:
  -> WooCommerce with dedicated DevOps, OR Shopify Plus B2B

Else if commerce logic IS the differentiator:
  -> Custom (Medusa, commercetools, or bespoke on Next.js + Stripe)

Else if you need React-grade front-end on Shopify catalog:
  -> Hydrogen or Next.js + Storefront API (only above ~$5M GMV)

This isn't gospel, but it correctly classifies about 90% of the briefs we see.

The Questions We Actually Ask in Discovery

When a client asks us to recommend a platform, we don't open a feature matrix. We ask:

• What's your current monthly order volume and 24-month projection?
• What's your gross margin per order, and how price-sensitive is platform fee as a line item?
• What systems do you need to integrate (ERP, WMS, PIM, CRM, accounting)? Name them.
• Who maintains the site at 2am when checkout breaks? In-house, agency retainer, or nobody?
• Are you selling B2C, B2B, or both? If both, do they share a catalogue?
• Which markets and currencies, and which payment methods are non-negotiable?
• What's your content-to-commerce ratio? Are you a brand that sells, or a shop with some content?
• What's the team's existing tech stack and hiring pipeline?

The answers usually point clearly at one option within 30 minutes. The rest of the discovery is just confirming the gut call with numbers.

Migration Reality Check

If you're already on a platform and considering moving, add a sober tax to your estimate. Migrations we've shipped tend to cost 1.5 – 2.5x the equivalent greenfield build once you factor in:

• Historical data migration (orders, customers, subscriptions, loyalty points)
• SEO preservation: URL mapping, redirects, structured data parity
• Re-integrating every tool that touched the old platform
• Re-training internal teams and rewriting SOPs
• A parallel-run period where both systems are live

A migration only pays back if the new platform unlocks revenue or saves operating cost that exceeds this tax within 18 – 24 months. "The admin is nicer" is not a business case.

Where We'd Start

If you're staring at this decision right now: don't start with a platform demo. Start with a one-page document listing your monthly orders, gross margin, top five required integrations, and who owns the site at 2am. Email it to two or three implementation partners (yes, including us) and see whether their recommendations converge.

If they all say the same thing, that's your answer. If they disagree, the disagreement itself is informative — usually it means your requirements aren't yet sharp enough to make the call, and another week of internal alignment will save you six months of regret.