Twelve brands. One platform. One AI layer.

The challenge

Twelve retail brands. Three different e-commerce platforms. No shared customer view, fragmented inventory, marketing operating in isolation per brand — and a CIO mandate to consolidate without disrupting peak-season trading.

The client had grown by acquisition. Each acquired brand kept its existing tech: four on Magento 1 (long out of support), six on WooCommerce of varying vintages, two on a custom PHP stack. Customer accounts didn't carry across brands. Orders couldn't fulfil from another brand's warehouse. Marketing teams couldn't see a customer who'd bought from three of the twelve brands.

Worse, Magento 1 was a security liability — and the client's previous re-platform attempt (vendor-led, big-bang cutover) had failed and been reverted, costing 14 months and a meaningful share of organic traffic. Trust in re-platforming was at its floor.

The approach

A single Shopify Plus organisation with twelve brand-specific storefronts on headless Hydrogen — sharing customer accounts, inventory, fulfilment, and a unified AI personalisation layer.

The architectural decision that unlocked everything: one Shopify Plus org, twelve sales channels. Each brand kept its identity through a Hydrogen-based storefront with brand-specific design, voice, and product catalogue. But underneath, customer accounts, gift cards, loyalty, inventory, and fulfilment all lived in one place.

We then layered a shared AI personalisation service across all twelve frontends — same recommendation engine, same search, same dynamic content — trained on the unified customer behaviour graph that became possible only after consolidation. A customer who browses Brand A and buys from Brand C now gets cross-brand recommendations that compound learning.

Migration was sequenced to remove all-or-nothing risk. We migrated in waves of three brands every six weeks, each wave running in parallel with the legacy system for two weeks before cutover. SEO redirect maps were the highest-priority artefact for every wave.

What we built

• One Shopify Plus organisation with 12 sales channels, shared customer accounts, gift cards, loyalty, and B2B portal
• 12 headless Hydrogen storefronts on Oxygen — edge-rendered globally, brand-specific design, mobile-first, 1.3s LCP target
• AI personalisation service — shared across all storefronts: recommendations, search, dynamic merchandising, A/B testing
• Snowflake data warehouse ingesting Shopify, ad platforms, fulfilment, returns, and POS — single source of truth for analytics
• SEO migration — full 301 redirect maps per brand, schema preservation, post-launch ranking monitoring with auto-alerting
• Order & inventory consolidation — cross-brand fulfilment, unified returns, single-warehouse view across brands
• Phased cutover — three brands per wave, staged over 18 weeks, with rollback path tested at every step

Tech stack

Commerce PlatformShopify Plus · 12 sales channels · B2B

StorefrontHydrogen · Remix · Oxygen edge runtime

AI PersonalisationCustom embeddings · Pinecone · GPT-4o for content

SearchAlgolia · vector search · query understanding

CloudAWS · Lambda · API Gateway · DynamoDB

Data PlatformSnowflake · dbt · Fivetran · Looker

Migration ToolingCustom catalogue ETL · order replay · redirect generator

MonitoringDatadog · Sentry · Shopify Flow · uptime SLO 99.95%

Outcomes

Honest reflections

The hardest part of this engagement wasn't technical. It was governance. Twelve brand teams, twelve points of view on what their storefront should do. We spent six weeks at the start getting agreement on the shared design system — what's tunable per brand and what's locked at the platform level. Without that document, the project would have shipped twelve increasingly different storefronts and lost the consolidation benefit.

The AI personalisation service was the most-questioned line item in the budget at kickoff. By month four, every brand team was negotiating for more model training time. The lesson is durable: shared AI services pay for themselves only after data consolidation — but once data is unified, AI becomes the highest-leverage investment available to multi-brand retail.

Capabilities used