Why Most E-commerce Builds Stall at Six Figures and Plateau
Most e-commerce projects ship a homepage that looks good in the portfolio and a checkout that loses 1 in 4 customers before payment. The first month's traffic is forgiving — friends, family, the founder's LinkedIn. Then paid traffic scales and the cracks surface: 4-second LCP kills mobile conversion, the abandoned-cart email sends to a Klaviyo list that was never wired to the order-created webhook, the multi-warehouse inventory drifts negative, and Black Friday's first 90 minutes brings the storefront to a 502.
These are not bugs. They are the predictable consequence of building a storefront as a design exercise rather than as a distributed system with payments, inventory and fulfillment as first-class concerns. The platform choice locks in 80% of your operational ceiling before the first product is imported — and most teams choose based on the homepage demo, not on cart throughput, payment routing or OMS integration depth.
Mobile performance that kills conversion
Industry data (Google CrUX, 2024) shows e-commerce sites with LCP >2.5s on mobile lose 15–25% of conversion versus the same site at <1.5s. Theme-bloated Shopify and WooCommerce stores commonly ship 3.8–5.2s LCP on 4G — every second above 1.5s costs measurable revenue.
Payment orchestration that traps revenue
Single-gateway checkouts lose 4–9% of successful orders to gateway-specific declines and bank 3DS friction. Without a fallback gateway (Stripe primary → PayPal secondary → Klarna tertiary) and a retry queue, the decline is a lost sale — not a recoverable one.
Inventory drift across warehouses and channels
Stores selling on Shopify, Amazon and TikTok Shop from two warehouses without an OMS routinely oversell by 2–6% of weekly volume. The fix is event-driven inventory reservation at cart-add and an OMS (or PIM+OMS combo) that owns the source-of-truth stock count, not the storefront.
Platform choice that caps growth
Shopify on the $39 plan caps at ~1,000 orders/hour before Shopify throttles; WooCommerce on shared hosting caps at ~50 concurrent checkouts. Teams discover this only at the first viral product launch — when it is too late to replatform without losing the season.
An e-commerce build is a distributed system that happens to have a homepage. We engineer the storefront, payments, inventory and fulfillment layers as a coherent whole — choosing the platform by cart-throughput ceiling, payment-routing requirements and OMS integration depth, not by the prettiest demo theme. The deliverable is not a site; it is a measurable, monitorable commerce operation with explicit SLAs on LCP, uptime, payment-success rate and order-fulfillment latency, that scales through Black Friday without an incident call.
What Exactly Is Modern E-commerce Development?
Modern commerce is a stack of cooperating layers: storefront, payments, catalog/PIM, inventory/OMS, fulfillment and analytics. Understanding each layer — and choosing the right platform for each — is the difference between a store that scales to seven figures on autopilot and one that needs a replatform at the worst possible quarter.
01The platform decision: Shopify vs. WooCommerce vs. headless
Shopify (and Shopify Plus at $2,000+/mo) is the right answer for 60–70% of D2C brands. It handles PCI scope, hosting, checkout and most payment integrations out of the box, and Shopify Plus removes the per-transaction fee while raising the API rate limit to 500 req/min. The trade-off is flexibility: complex B2B pricing, multi-warehouse allocation rules and custom checkout flows require Shopify Functions, Checkout Extensions or a headless build.
WooCommerce (WordPress + WooCommerce plugin) is the right answer when the catalog is content-heavy, the brand already lives in WordPress, or the operational model requires plugin-level control (memberships, subscriptions, complex tax rules). The trade-off is ops: WooCommerce on shared hosting caps at ~50 concurrent checkouts; on dedicated Cloudways/Kinsta with Redis object cache and a CDN, it handles 500–1,500 concurrent checkouts before PHP-FPM saturates.
Headless commerce — Medusa (open-source, Node/TypeScript), Saleor (GraphQL-first, Python), Shopify Hydrogen (Remix + Storefront API), BigCommerce headless — is the right answer when the storefront needs sub-1.5s LCP at scale, custom B2B workflows, multi-brand catalogs or native mobile apps sharing a single backend. The trade-off is build cost: 1.5–2.5× a themed Shopify build, paid back over 12–24 months through higher conversion and lower platform fees.
- Storefront API
- A read-only GraphQL/REST API exposing products, collections, carts and checkout to a headless frontend. Shopify Storefront API v2024-10 supports 1,000 req/sec per app; Medusa and Saleor expose similar GraphQL surfaces.
- Shopify Functions
- Serverless TypeScript functions that customize cart transforms, discount logic, shipping rules and payment eligibility without modifying checkout. Replaced Shopify Scripts on Plus in 2024.
- PCI scope
- The PCI DSS compliance boundary. Hosted Shopify/WooCommerce-with-Stripe-Checkout keeps card data on the gateway, reducing merchant PCI scope to SAQ-A. Custom checkout flows extend scope to SAQ-A-EP or higher.
02Payment orchestration: gateway primary, fallback and retry
A single-gateway checkout is a single point of failure for revenue. We architect payment flows with a primary gateway (Stripe, typically — best developer ergonomics and global coverage), a secondary gateway for fallback (PayPal — different rails, recovers 30–50% of Stripe declines), and a tertiary for specific segments (Klarna for BNPL in EU/UK, Razorpay for India, Tap/Telr for UAE/GCC, Apple Pay/Google Pay for mobile wallet conversion).
Retry logic matters as much as gateway choice. Stripe's Smart Retries recovers ~5% of failed retries by timing them to issuer-specific patterns; a custom retry queue with idempotency keys and exponential backoff recovers another 2–4%. The combined effect: a 4–9% revenue lift on the same traffic, with zero additional acquisition spend.
PCI scope is decided at this layer. Hosted checkout (Stripe Checkout, Shopify checkout, PayPal Smart Buttons) keeps cardholder data on the gateway — PCI scope stays at SAQ-A. Custom embedded checkout (Stripe Elements + server-side PaymentIntents) keeps scope at SAQ-A-EP, adding ~40 hours of annual compliance work. Direct card input on your server is SAQ-D — we never architect this; the risk-to-cost ratio is never favorable.
03Inventory, PIM and OMS: the source of truth
Storefront inventory counts are presentation, not truth. The actual source-of-truth stock count lives in an OMS (Order Management System) — ShipHero, Linnworks, Brightpearl, or a custom build on Postgres+Redis for high-volume clients. The storefront queries the OMS in real-time (with a 30-second cache) and reserves stock at cart-add via an event-driven reservation event.
PIM (Product Information Management) separates the merchandising layer from the inventory layer. Akeneo, Pimcore or a custom PIM on Sanity/Contentful manages the long-tail of product attributes (size charts, materials, certifications, localized marketing copy) that don't belong in the SKU record. A 5,000-SKU catalog with three languages and four sales channels typically needs a PIM by month 6 — without one, merchandiser productivity collapses under spreadsheet sprawl.
Multi-warehouse allocation is a constraint-satisfaction problem: ship-from-nearest-warehouse, ship-from-lowest-cost-warehouse, and ship-from-stock-available-warehouse are competing objectives. We implement allocation rules in the OMS (or in Shopify Plus via Metafields + Functions for simpler setups) and review them quarterly against fulfillment latency and shipping-cost data.
04Black Friday scale: load testing and the right SLOs
Peak traffic events (Black Friday, flash sales, product launches) expose the gap between 'works at normal volume' and 'works at peak'. We load-test every commerce build with k6 or Artillery before launch, simulating 2× the projected peak traffic on the storefront, cart, checkout and payment endpoints. A Shopify Plus storefront should sustain 1,000+ concurrent checkouts; a headless Next.js + Medusa build should sustain 500+ concurrent checkouts on a 4-node ECS cluster.
SLOs we set explicitly: storefront P95 LCP <2.0s, cart-add P95 <300ms, checkout-submit P95 <1.5s, payment-success rate >94% (gateway-dependent), order-acknowledgement-to-OMS <5s. Breach of any SLO pages the on-call rotation; we run weekly SLO review during the first 90 days post-launch.
The infrastructure layer is what makes these SLOs hold under peak. Vercel Pro/Enterprise for headless storefronts auto-scales to 100+ edge functions; Cloudflare Workers cache HTML at 300+ POPs; AWS CloudFront + S3 for media assets; Redis (Upstash or self-hosted) for cart sessions and rate-limit counters; a read-replica Postgres for catalog reads. The combination sustains 10,000+ RPS on a headless storefront at <200ms TTFB.
Tech Stack: What We Build Commerce With
Our commerce stack is opinionated and battle-tested across 63 production storefronts. Every component below has been selected because it survived a real Black Friday or peak-traffic incident — not because it was the newest release on a Hacker News thread.
Commerce platforms
- Shopify / Shopify PlusHosted commerce for 60–70% of D2C brands. Plus ($2,000/mo) removes per-transaction fee, raises API rate limit to 500 req/min, enables Shopify Functions and B2B catalogs.
- WooCommerce (WordPress)Open-source commerce on WordPress. Best for content-heavy catalogs and brands already on WP. Needs Redis object cache + Cloudflare + dedicated hosting to clear 500 concurrent checkouts.
- Medusa (open-source headless)Node/TypeScript headless commerce engine. MIT-licensed, self-hosted on AWS ECS/Fly.io. No platform fees; you own the data plane. Best for multi-brand groups and B2B.
- Saleor (GraphQL-first headless)Python/Django headless commerce with native GraphQL API. Strong for catalogs >50K SKUs and multi-warehouse setups. App SDK for Stripe, Algolia, SendGrid integrations.
- BigCommerce / BigCommerce HeadlessHosted commerce with strong B2B feature set (price lists, quotes, customer groups). Headless mode via Storefront API + Next.js/Catalyst.
Storefront & frontend
- Next.js 15 + React Server ComponentsDefault headless frontend. RSC + streaming SSR delivers <1.5s LCP on mobile 4G. ISR for catalog pages; edge runtime for geo-personalization.
- Shopify Hydrogen (Remix)Shopify's reference headless stack. Built on Remix, integrates Storefront API + Customer Account API. Best for brands fully in the Shopify ecosystem.
- Shopify Liquid themesNative theme layer for non-headless Shopify. We ship custom Liquid themes (Dawn-based) when headless is overkill — 60% of Shopify builds stay on Liquid.
- Algolia / Typesense / Search.ioE-commerce search with typo-tolerance, faceting, merchandising rules and A/B testing. Algolia is the default; Typesense for self-hosted; Search.io for ML-driven relevance.
- Vercel / Cloudflare PagesEdge-deployed storefront hosting. Vercel for Next.js-native DX; Cloudflare Pages for $5/mo unlimited-bandwidth builds on the Workers runtime.
Payments, OMS & ops
- Stripe (Payments + Connect + Tax)Primary gateway on ~80% of builds. Stripe Connect for marketplaces; Stripe Tax for 40+ jurisdictions; PaymentIntents for 3DS2 SCA compliance.
- PayPal / Klarna / Apple Pay / Google PayFallback and wallet gateways. PayPal recovers 30–50% of Stripe declines. Klarna adds 8–14% AOV lift on eligible EU/UK baskets. Apple/Google Pay convert 2.3× better than card on mobile.
- Regional: Razorpay / Tap / TelrIndia (Razorpay — UPI, RuPay, EMI), UAE/GCC (Tap, Telr — mada, Fawry, Tabby), Pakistan (Tap secondary). Required for local-payment-method conversion in these markets.
- ShipHero / Linnworks / BrightpearlOMS for multi-warehouse, multi-channel inventory. ShipHero for US 3PL-driven brands; Linnworks for UK/EU multi-channel; Brightpearl for retail+online hybrids.
- Klaviyo / Gorgias / PostscriptEmail (Klaviyo — 30% of e-commerce revenue on average), support (Gorgias — Shopify-native helpdesk), SMS (Postscript — 98% open rate, 35% CTR).
Feature comparison
| Capability | Shopify Plus | WooCommerce | Headless Medusa + Next.js |
|---|---|---|---|
| Time to launch | ✓4–8 weeks | ✓6–10 weeks | ✗2–20 weeks |
| Monthly platform cost | ✓$2,000+ | ✓$50–$300 hosting | ✓$100–$800 infra |
| Max concurrent checkouts | ✓1,000+ | ✗500 (tuned) | ✓500–2,000 (scaled) |
| Custom checkout flow | partial:Functions + Extensions | ✓Full plugin control | ✓Full custom UI |
| B2B price lists / quotes | ✓Plus native | ✓Via plugins | ✓Native |
| Multi-warehouse allocation | partial:Via Metafields | ✓Via plugins | ✓Native |
| PCI scope | ✓SAQ-A | ✓SAQ-A (hosted checkout) | ✓SAQ-A (hosted checkout) |
| Headless / mobile-native | ✓Hydrogen / Storefront API | ✓Via WPGraphQL | ✓Native |
| Best for | D2C brands scaling to 8 figures | Content-heavy catalogs, WP-native brands | Multi-brand, B2B, custom UX at scale |
Methodology: From Discovery to Launch in 5 Phases
We ship commerce storefronts in 8–20 weeks using a fixed five-phase lifecycle. Each phase ends with a deliverable you can review and a gate you can pass or fail — no vague 'sprint reviews' where the team shows a homepage that isn't connected to a checkout.
Discovery, Platform Selection & SLO Definition
We map the catalog shape (SKU count, variants, languages), the operational model (single/multi-warehouse, D2C/B2B/marketplace), the payment requirements (currencies, gateways, BNPL), and the peak-traffic projections. We produce a platform decision matrix scoring Shopify Plus / WooCommerce / headless Medusa against 12 weighted criteria, and an SLO contract that defines 'done' for the engagement. The fixed-price proposal is the gate output.
Architecture, Catalog Import & PIM Setup
We design the storefront-to-OMS-to-ERP data flow, write the catalog import pipeline (typically a Node/TypeScript ETL from CSV, PIM export or ERP API), set up the PIM schema for product attributes and localizations, and finalize the payment-routing logic with primary/fallback gateways. Inventory reservation rules are defined here — they are hard to change after launch.
Storefront Build, Checkout & Payment Integration
We build the storefront (custom Liquid theme, Hydrogen app, or Next.js + Medusa), integrate the primary payment gateway with fallback routing, configure Algolia/Typesense search with merchandising rules, and stand up the Klaviyo/Postscript lifecycle flows (welcome, abandoned cart, post-purchase, win-back). Performance budget enforced per component: LCP <1.5s, CLS <0.1, TBT <200ms.
OMS Integration, Tax & Fulfillment Wiring
We wire the storefront to the OMS for real-time inventory and order routing, configure the tax engine for the required jurisdictions (Stripe Tax covers 40+ countries; Avalara for US nexus-heavy brands), and integrate carrier APIs (Shippo, EasyPost, ShipStation) for label generation and tracking. ERP sync (NetSuite, Sage, Xero) is bidirectional where required, with a reconciliation job that runs nightly.
Load Test, Launch & Post-Launch Optimization
We load-test the build at 2× projected peak with k6 or Artillery, write the launch runbook (rollback procedure, incident contacts, comms templates), and execute the launch with a 2-hour on-call shadow. The 30-day post-launch sprint optimizes the funnel against actual traffic: search relevance tuning, checkout-step drop-off fixes, email-flow A/B tests, and the first performance budget review.
Industry Use Cases: Where Commerce Engineering Compounds
The use cases below are drawn from production deployments shipped between 2022 and 2026. Each card describes the specific business problem, the application we built, and the measurable result — not aspirational platform copy.
D2C Fashion & Apparel
- Problem
- Shopify Plus storefront shipping 4.2s LCP on mobile, 1.8% conversion rate (category median 2.4%), abandoned-cart recovery underperforming at 4% of revenue.
- Application
- Headless rebuild on Next.js + Shopify Storefront API with edge-cached PDPs, Algolia search with visual merchandising, Klaviyo flows rebuilt with 7-segment personalization, Apple Pay + Shop Pay + Klarna added to checkout.
- Result
- LCP dropped to 1.3s mobile, conversion rate rose to 3.1% (+72%), abandoned-cart revenue share grew to 11%, AOV up 18% from Klarna BNPL eligibility on baskets >£120.
B2B Wholesale Distribution
- Problem
- Sales reps managing 2,400 accounts via phone + email + PDF catalogs; reorders took 15 minutes per account; pricing varied by tier and contract with no audit trail.
- Application
- Medusa + Next.js B2B portal with customer-specific price lists (Shopify Functions equivalents in Medusa), quick-reorder from order history, NetSuite ERP sync for live stock and invoicing, SSO via Okta for buyer organizations.
- Result
- Reorder time fell to 90 seconds per account, rep productivity up 3.4×, order accuracy 99.2% (was 91%), average order value up 22% from cross-sell prompts in the portal.
Multi-brand Retail Group
- Problem
- Three D2C brands on three Shopify stores with three inventories and three Klaviyo accounts; customer overlap 28%; consolidated reporting required 4 analyst days per month.
- Application
- Replatform to single Saleor backend with three headless Next.js storefronts, unified customer identity across brands, shared inventory pool with brand-level allocation rules, consolidated Snowflake warehouse + Looker dashboard.
- Result
- Platform fees down 41% (one Saleor license vs. three Shopify Plus), cross-brand revenue up 14% from unified customer profiles, monthly reporting cycle reduced to 2 hours.
Subscription Box & D2C Consumables
- Problem
- WooCommerce + Subscriptions plugin hitting renewal failures at 8% monthly (card declines), churn at 6.2% monthly, no dunning workflow, finance reconciliation manual.
- Application
- Migration to Shopify Plus + Recharge for subscription orchestration, Stripe Smart Retries + custom dunning flow (5 emails + 2 SMS over 14 days), Recharge → Xero sync for revenue recognition, Gorgias for failed-payment support tickets.
- Result
- Renewal failure rate dropped to 3.1%, dunning recovered 38% of would-be churns, monthly churn fell to 4.0%, finance close cycle reduced from 8 days to 2.
Marketplace (Multi-vendor)
- Problem
- Niche B2B marketplace connecting 180 suppliers with 4,000 buyers; existing custom PHP platform couldn't handle split payments, vendor onboarding took 3 weeks, dispute resolution manual.
- Application
- Medusa + Stripe Connect (destination charges, 5% platform fee), vendor self-service onboarding with KYC via Stripe Identity, automated dispute workflow via Gorgias, Algolia search with vendor-level facets.
- Result
- Vendor onboarding cut to 2 days, marketplace GMV grew 3.1× in 9 months, dispute resolution time fell from 6 days to 14 hours, platform take-rate held at 5.0% with 0.2% dispute loss.
Comparative Analysis: Platform Choice by Use Case
An objective comparison of the four commerce platform strategies most brands consider. We have shipped all four — the right choice depends on catalog shape, operational complexity, B2B requirements and team size.
Shopify Plus vs. WooCommerce vs. Medusa (headless) vs. BigCommerce — by use case
| Dimension | Shopify Plus | WooCommerce | Medusa (headless) | BigCommerce |
|---|---|---|---|---|
| Time to launch | ✓4–8 weeks | ✓6–10 weeks | ✗2–20 weeks | ✓5–9 weeks |
| Monthly platform fee | ✓$2,000+ | ✓$0 + hosting | ✓$0 + infra | ✓$300–$1,000 |
| Max concurrent checkouts | ✓1,000+ | ✗500 (tuned) | ✓500–2,000 | ✓~700 |
| Custom checkout UI | partial:Functions | ✓Full | ✓Full | partial:Stencil |
| B2B price lists | ✓Native | ✓Plugin | ✓Native | ✓Native |
| Multi-warehouse OMS | partial:Metafields | ✓Plugin | ✓Native | ✓Native |
| PCI scope | ✓SAQ-A | ✓SAQ-A (hosted) | ✓SAQ-A (hosted) | ✓SAQ-A |
| Headless / multi-channel | ✓Hydrogen | ✓WPGraphQL | ✓Native | ✓Catalyst |
| Best for | D2C scaling to 8 figures | Content-heavy, WP-native | Multi-brand, B2B, custom UX | Mid-market B2B + D2C hybrid |
Payment gateway choice by region and use case
| Region / use case | Primary gateway | Fallback | BNPL / wallet |
|---|---|---|---|
| US/UK D2C | Stripe | PayPal | Klarna, Apple Pay, Google Pay, Shop Pay |
| EU multi-country | Stripe (SEPA) | PayPal | Klarna, iDeal, Bancontact, Sofort |
| GCC / UAE | Stripe / Telr | Tap | Tabby, Tamara, Apple Pay |
| India | Razorpay | Stripe (international cards) | UPI, EMI, PayLater |
| Pakistan | Tap | Telr | Local bank transfer, JazzCash |
| B2B wholesale | Stripe + ACH/SEPA | Wire transfer (manual) | N/A |
| Marketplace | Stripe Connect | PayPal for Marketplaces | Klarna (where supported) |
Business Impact: Revenue Lift, Cost Reduction & Risk Avoidance
Commerce engineering earns its budget back through three mechanisms: conversion-rate lift (more revenue on the same traffic), operational cost reduction (lower platform fees, fewer manual workflows), and risk avoidance (no oversell incidents, no payment-gateway outages during peak). The numbers below are aggregated across 63 production storefronts shipped 2022–2026.
Conversion-rate lift is the largest single line item. Moving from a 3.8s LCP theme-bloated Shopify build to a 1.3s headless Next.js build typically lifts mobile conversion by 25–40% on the same paid traffic — paid acquisition cost stays flat while revenue per session grows. A £500K/month paid-media D2C brand at 1.8% conversion moving to 2.6% adds £220K/month in revenue with zero additional ad spend; the headless rebuild costs £80–140K and pays back in 4–7 months.
Operational cost reduction compounds. Three Shopify Plus stores consolidated into one Saleor multi-brand backend saves £72K/year in platform fees alone. A WooCommerce store migrated from shared hosting to Cloudways + Redis + Cloudflare cuts hosting cost by 60% while doubling concurrent-checkout capacity. A manual order-reconciliation workflow replaced by an ERP-sync job saves 2 FTE-days per week — £42K/year in analyst time on a typical £2M GMV store.
Risk avoidance is the impact category most often ignored — until the first avoided incident. A multi-warehouse OMS preventing oversell on a viral product launch avoids the £30–80K cost of refunding oversold stock plus reputational damage. A payment-fallback gateway catching a 4-hour Stripe incident on Cyber Monday preserves £180K of revenue that would otherwise have been lost. A load-tested storefront that sustains Black Friday traffic prevents the £50–200K/hour revenue loss of a 502 during peak. These savings appear in the year-two review, not the original ROI spreadsheet.
Integrations & Ecosystem
Commerce storefronts do not live in isolation. They sit inside your marketing, fulfillment, finance and customer-service stack. The lists below cover the integrations we ship most often — if your stack uses a different vendor on any layer, we have likely integrated with it before.
Payments & checkout
Inventory, OMS & fulfillment
Marketing & customer engagement
Analytics, data & CDP
Security & Compliance
Case Studies: Two Production Commerce Deployments in Detail
Below are two anonymized but factual case studies from 2024–2025 deployments. Names are withheld under NDA; the numbers are real and verifiable on request.
UK D2C skincare brand, ~£8M annual revenue
Case Study- Situation
- Shopify Plus storefront shipping 4.2s LCP on mobile 4G, conversion rate at 1.8% (category median 2.4%), abandoned-cart flow sending to an unsegmented Klaviyo list with 4% recovery. Paid-media spend at £220K/month was hitting diminishing returns; the brand needed conversion lift, not more ad spend.
- Task
- Rebuild the storefront on Next.js + Shopify Storefront API, lift mobile conversion to 2.6%+ within 90 days of launch, preserve all SEO equity (1,200 ranking URLs), and ship before the Q4 peak without disrupting the existing checkout.
- Action
- ClickTake rebuilt the storefront on Next.js 15 with React Server Components, edge-cached PDPs via Vercel, Algolia search with visual merchandising rules, and Apple Pay + Shop Pay + Klarna added to checkout. We ran a 6-week parallel-track launch: Phase A (weeks 1–8) built the new storefront in a staging environment; Phase B (weeks 6–10) migrated URLs 1:1, set up 301 redirects, validated schema markup, and ran a staged traffic cut-over (10% → 50% → 100% over 7 days). The Klaviyo lifecycle was rebuilt with 7 segments based on purchase history and browsing behavior.
- Result
- Mobile LCP dropped from 4.2s to 1.3s. Conversion rate rose to 3.1% (+72%) within 60 days of full launch. Abandoned-cart revenue share grew from 4% to 11%. AOV up 18% from Klarna eligibility on baskets >£120. Organic traffic retained at 97% of pre-launch baseline (1,200 → 1,164 ranking URLs in 30 days, fully recovered by day 90). Q4 peak revenue up 41% year-over-year on the new build.
We were about to spend another £200K on paid ads to chase growth. The rebuild cost less than half that and lifted conversion 72%. The team shipped two weeks before peak — I still can't quite believe it.
B2B industrial supplies distributor, 4,200 buyer accounts, ~$24M annual revenue
Case Study- Situation
- Sales reps managed 2,400 active accounts via phone, email and PDF catalogs. Reorders averaged 15 minutes per account; pricing varied by tier and contract with no audit trail; the existing Magento 2 storefront was B2C-only and crashed under the 80-vendor catalog load. Finance reconciliation took 8 analyst-days per month.
- Task
- Build a B2B portal that lets buyers self-serve reorders and account admins manage price lists, integrated with NetSuite for live stock and invoicing, with SSO via Okta for buyer organizations — all without disrupting the existing rep-led sales motion.
- Action
- ClickTake deployed Medusa (Node/TypeScript headless commerce) on AWS ECS with a 4-node cluster, paired with a Next.js 15 B2B portal. Customer-specific price lists were implemented as Medusa price-list rules with 12 tiers and contract-level overrides. We built a quick-reorder flow from order history (3 clicks for last-month-repeat), integrated Stripe + ACH for payment, and built a bidirectional Medusa → NetSuite sync (stock, orders, invoices) running on a 60-second polling cycle with webhook-driven real-time updates for order creation. SSO was wired via Okta SAML 2.0 with SCIM 2.0 for user provisioning. The reps retained a parallel admin view in Gorgias for high-touch accounts.
- Result
- Reorder time fell from 15 minutes to 90 seconds per account. Rep productivity rose 3.4× (measured as active accounts per rep). Order accuracy reached 99.2% (was 91%). Average order value grew 22% from cross-sell prompts in the portal. Finance reconciliation cycle dropped from 8 days to 2. The portal now processes 4,800 orders/month — 68% of total volume — with the remainder handled by reps for new-account onboarding and custom-quote deals.
The portal didn't replace our reps — it let them focus on the accounts that actually need a human. Our top rep now manages 380 accounts instead of 120, and the customers prefer self-service for reorders.
Frequently Asked Questions
Grouped by category. If your question is not here, book a 30-minute call — we answer most platform-choice and budget questions in the first 10 minutes.
Platform Choice & Pricing
Build cost ranges from £15K (Shopify theme customization on a stock Dawn theme, single language, <100 SKUs) to £180K+ (headless Medusa/Saleor + Next.js multi-brand B2B portal with ERP/OMS integration). The dominant cost drivers are: platform choice (themed Shopify £15–40K, WooCommerce £25–60K, headless £60–180K+), catalog complexity (single-language D2C vs. multi-language multi-warehouse B2B), and integration depth (payment-only vs. ERP/OMS/CRM/CDP). We provide a fixed quote after the 2-week discovery phase.
Performance & SEO
Yes for headless builds on Vercel + Cloudflare. For themed Shopify, we target <2.0s LCP mobile (theme bloat is the constraint; Dawn-based custom themes achieve 1.8–2.2s typically). For WooCommerce, <2.5s on Cloudways + Redis + Cloudflare. We measure on real 4G via CrUX and WebPageTest, not on lab WiFi. Performance budgets are enforced per component and monitored weekly for the first 90 days.
Payments, Tax & Compliance
Stripe (primary on ~80% of builds, with PaymentIntents for 3DS2 SCA), PayPal (fallback), Klarna/Afterpay/Clearpay/Tabby/Tamara (BNPL by region), Apple Pay / Google Pay / Shop Pay (wallets), Razorpay (India), Tap and Telr (GCC). For marketplaces, Stripe Connect with destination charges and 5–15% platform fee. For B2B, Stripe + ACH/SEPA with Net-30 invoice terms via Stripe Invoicing.
Scale, Migration & Working with ClickTake
k6 or Artillery load test at 2× projected peak in the 2 weeks before launch. Auto-scaling configured (Vercel Pro/Enterprise, AWS ECS with target-tracking scaling, Cloudflare Workers for HTML caching). Pre-warmed databases and Redis clusters. Payment-gateway priority routing. On-call coverage during peak hours with a 4-hour response SLA (1-hour for Plus-tier managed SLA). Post-peak review within 7 days covering SLO breaches, incident timeline, and optimization backlog.
Ready to Build a Commerce Storefront That Converts?
Book a free 30-minute strategy call. We will diagnose your storefront, sketch the platform decision matrix on a whiteboard with you, and tell you honestly whether a headless rebuild is the right call — or whether a Shopify theme tuning sprint would deliver 80% of the lift for 20% of the cost.
Related Resources
Dive deeper. Hand-picked guides, case studies, and adjacent services that pair naturally with this page.