Before you build: the decisions that compound

Most failed e-commerce launches fail before any code is written. The decisions made in the first two weeks shape the next two years.

Three questions to answer before you choose a platform:

  1. What are you actually selling, and to whom? "Anyone interested in [category]" is not an answer. The first 100 customers come from a tight, specific audience. Define them.
  2. What does the unit economics math actually look like? Cost of goods, payment processing (typically 2.5–3.4%), shipping (varies wildly), agency or marketing budget, refund rate. If gross margin per order is under 35–40%, paid acquisition rarely pencils out.
  3. What's the product depth — one SKU or many? Single-SKU brands (one hero product, color/size variants) have a completely different operating profile from multi-SKU catalogues. The platform decisions diverge.

Step 1: Choose your platform

The decision tree we use with clients:

| Your situation | Platform | Why | |---|---|---| | First store, <500 SKUs, single market, want fastest launch | Shopify Basic or Advanced | Lowest operational overhead, best app ecosystem, default-good defaults | | 500–5,000 SKUs, multi-region, multi-currency | Shopify Plus | B2B pricing tiers, multi-store, checkout customization | | Unusual catalogue logic, B2B with negotiated pricing, custom checkout flows | Medusa.js (headless, open-source) | Owns the database, no platform lock-in, infinitely customizable | | Enterprise, multiple brands, complex global tax | Commerce Layer + custom Next.js or Shopify Plus with B2B | Multi-storefront, multi-region, full ownership | | Selling digital products or services | Stripe Checkout, Paddle, or Lemon Squeezy | Lighter, faster, no inventory management |

We do not recommend WooCommerce, Magento (Adobe Commerce open-source), BigCommerce, or Wix for new builds in 2026. Each has its constituency; the ROI math doesn't pencil out against Shopify for most cases.

For more on each: see /online-store-development.

Step 2: Set up the foundation

Once the platform is chosen:

Legal and operational

  • Business entity — typically LLC (US/Canada) or Limited Company (UK/EU equivalent). Single-member LLCs work for most solo founders.
  • Bank account in the business name.
  • Sales tax / VAT registration in your home jurisdiction. International sales tax (US economic nexus, EU OSS scheme, UK VAT post-Brexit) is more complex — we cover this in detail with clients at engagement.
  • Terms of service, privacy policy, refund policy. Templates are fine to start; have a lawyer review before significant revenue.

Payment processing

  • Stripe for cards in most markets; rate is typically 2.9% + €0.30.
  • PayPal as a secondary because some customers still strongly prefer it (especially older demographics and certain EU markets).
  • Apple Pay and Google Pay enabled — they materially improve mobile conversion.
  • Klarna, Affirm, or Afterpay if your AOV is over €80–€100 and your margins support it.

Shipping setup

  • Carrier accounts. Canada Post, DHL, FedEx, UPS, USPS, regional carriers — set up business accounts so you get rate negotiation eligibility from volume.
  • Real-time rates at checkout. Static shipping is fine to launch with; real-time rates improve conversion and reduce loss on under-quoted shipping.
  • Free-shipping threshold. Most stores benefit from one. Set it slightly above your current AOV.

Taxes

  • Stripe Tax is the simplest path for most stores.
  • TaxJar or Avalara for higher complexity (heavy US state-by-state, EU VAT OSS, etc.).
  • Built into Shopify Tax for Shopify stores.

Step 3: Build the catalogue

The mistake most first-time stores make: they spend three weeks on the homepage and three days on the product pages. Reverse this. Customers convert on product pages.

For each product:

  • High-quality photography — minimum 5 angles, ideally with lifestyle context. Phone photography is fine; just respect the basics (good lighting, neutral background, consistent style).
  • Clear, accurate, scannable copy. Short paragraphs, bullet-pointed features, an honest description of who it's for and who it isn't.
  • Real reviews as they accumulate. The first 5–10 reviews are the hardest; do not buy them.
  • Schema markup — Product, Offer, AggregateRating (when you have reviews). Shopify does this by default for most themes; verify with Google's Rich Results Test.
  • Cross-sells and accessories below the fold or on the cart page.

Step 4: Set up analytics and tracking

Before launch, not after:

  • Google Analytics 4 — see our GA4 setup guide.
  • Meta Pixel + Conversions API if you'll run any Meta ads.
  • Google Ads Tag + Enhanced Conversions if you'll run any Google ads.
  • Klaviyo or Shopify Email for transactional and marketing email — set up abandoned cart, browse abandonment, post-purchase, and review-request flows.
  • Heatmap and session recording — Microsoft Clarity (free) is fine to start.

Step 5: Launch quietly first

Soft launch to a small audience (friends, family, an existing email list, your own social following) for 2–4 weeks before paid acquisition or aggressive PR. The goals of the soft launch:

  1. Catch the embarrassing bugs.
  2. Validate the checkout actually works for real customers.
  3. Get your first 5–15 reviews.
  4. Pressure-test your customer service capacity.
  5. Identify the questions customers actually ask (which becomes your FAQ content).

Step 6: The first 90 days

In priority order:

  1. Talk to every customer. Email each one after their order, ask why they bought, what almost stopped them, what they'd change. This is the highest-ROI marketing research you can do.
  2. Iterate on product pages based on what you hear.
  3. Build SEO foundation — see SEO for business: where to start.
  4. Set up one paid acquisition channel properly before adding a second. Meta or Google, not both at week one. €1,500–€3,000/month is enough to learn whether it works.
  5. Email marketing — abandoned cart, post-purchase, win-back. These flows produce 20–35% of e-commerce revenue when set up properly.
  6. Reviews — make collection systematic. SMS or email 5–7 days post-delivery with a one-click review link.

What kills early stores

Patterns we see:

  • Over-budgeting on design, under-budgeting on product photography. A €500 logo with €50 product photos converts worse than the reverse.
  • Building for a hundred features at launch. Launch with the basics; add features when customer behavior tells you what to add.
  • Spending paid acquisition before organic foundation. Paid traffic to a low-converting site is expensive learning.
  • Ignoring email until "we have more list." The opposite is correct — email starts working at 50 subscribers, not 5,000.
  • Trying to be on every channel. Pick one acquisition channel and one social channel. Add the second after the first is profitable.

Related

For the full agency-grade build: /online-store-development. For the marketing strategy that surrounds it: /marketing-strategy-development. For paid acquisition: /targeting-ads. For SEO: /seo-promotion.

Written by the Marketing MIX commerce team. Last reviewed: 2026-05-13.