Track A — AI engineering

Running a multi-£m Shopify Plus store on Claude Code

This is the case study we point at when prospects ask whether AI-native ecommerce operations actually work, or whether it’s a pitch deck waiting to fall over in production.We run a multi-£m UK Shopify Plus store with 500,000 customers across the catalogue. The entire store is managed through a custom Claude Code stack we built — MCP servers connecting Claude to Shopify Admin GraphQL, GSC, Ahrefs and our own internal data, custom Claude Code skills authored for every recurring workflow, sub-agents for specific high-frequency decisions. The result: half the operations team we used to need, hundreds of hours saved every month, and a catalogue of 7,000+ variants kept tighter than most stores at five times the size.The same stack we deploy to clients in our AI engineering engagements was built here first, on our own store, as a daily working tool.

The store

A consumer retail Shopify Plus store. Multi-£m annual revenue. 500,000+ customer accounts. Catalogue of around 2,000 products and 7,000+ variants when you count every flavour, size and configuration. Active SKU lifecycle management — new launches monthly, clearances quarterly, supplier changes constantly. Mix of own-brand and reseller products. UK-focused but ships internationally.The store has been running for years. The AI-engineering layer on top of it is more recent — the result of building, breaking and rebuilding a stack of tools that turned the daily ops grind into something tractable for a small team.

The brief — to ourselves

We didn’t get hired for this. We’re the operators. The brief evolved over time but settled on three things we wanted from the AI tooling:One — replace the merchandising grind. Variant-level decisions across thousands of SKUs every quarter were eating most of the team’s week. Clearance triage, restock prioritisation, slow-mover identification, attribute normalisation. All of it doable by a person, none of it valuable to do by a person.Two — make content scale linearly with effort, not headcount. SEO needs collection content, product descriptions, FAQ blocks, meta titles and descriptions across hundreds of pages. Doing that by hand limits how often you can refresh, which limits ranking. Doing it with a generic AI tool makes it sound generic. We needed something brand-tuned that produces content we’d actually publish.Three — own the tooling. Plenty of SaaS tools claim parts of this, but they all hit a wall when you want them to do something specific to your store. We wanted code we owned, prompts we tuned, and workflows that fit our specific catalogue rather than the lowest common denominator.What we built does all three. Below.

What we built

Catalogue operations engine

A Python agent against the Shopify Admin GraphQL API, used for any workflow involving doing the same thing to hundreds or thousands of variants. First job was a cost-update run — 7,000+ variants updated in one bulk operation, 83% coverage on active products. Since then it’s done bulk metafield writes (collection content blocks, FAQ metaobjects, content top blocks), bulk meta title and description rewrites, redirect imports, attribute normalisation, and product creation from supplier spec sheets.If a workflow involves the same operation across the catalogue, this is the engine.

Variant-level merchandising agent

Variant-aware clearance and lifecycle decisions. Pulls 90-day sales velocity per variant, applies a rule that took some painful learning to land on — only clearance-tag a product if every variant is slow, cut only the slow variants if the product is mixed — and writes the changes back through GraphQL.The first full run did 248 SKU lifecycle decisions in a single day, variant-aware: 47 hardware, 201 e-liquid. Replaces roughly 40 hours of manual triage per week when run regularly.

SEO content sub-agents

Custom Claude Code agents for collection page rewrites, product descriptions, FAQ generation and meta titles. Each one is brand-voice tuned, fed structured input from Ahrefs and GSC, and writes output that meets our internal SEO checklist — primary keyword in H1 and first paragraph, semantic variations woven in, FAQ blocks where they make sense, internal links to sibling collections.We’ve used them across 200+ collections on our own store. The output isn’t perfect — nothing AI writes is — but the editing time is roughly 80% less than writing from scratch, and the quality bar is high enough that the published content wouldn’t read as AI-generated to a careful reader.

Custom MCP servers across the stack

In-house MCP servers connecting Claude to:

Shopify Admin GraphQL (read and write — variants, metafields, metaobjects, redirects, bulk operations)
Ahrefs (keyword data, ranking, SERP overview, competitor research)
Google Search Console (keywords, pages, performance, anonymous queries)
Sanity (for content from the property finance build)
Internal merchandising data store (sales velocity, stock levels, supplier data)

These let Claude do real work on real systems. Not chat about the systems. The MCP layer is the difference between AI tooling that helps and AI tooling that runs the operation.

Claude Code skills library

Authored skills for every recurring workflow:

Collection creation and content upload
Shopify product listing imports (turn pasted supplier data into properly listed products with metafields, variants, SEO populated)
Weekly merchandiser reporting
Clearance triage workflows
Content writing for collection and product pages
Competitor analysis
SEO content optimisation

Skills turn ambiguous requests into deterministic workflows. “List these 12 products” goes from a 30-minute conversation about the brief to a 2-minute call to a skill that knows the brief.

Weekly merchandiser report agent

Produces a weekly report covering revenue by collection, top movers, slow movers needing attention, cost-margin alerts, low-stock warnings, and content opportunities surfaced from GSC. Pulls from Shopify, GSC and our internal store. The agent does what a junior merchandiser used to spend a day on, in roughly 10 minutes of compute, every Monday. Output is a Markdown document that goes straight into the team channel.

Results

The honest numbers:Team size cut roughly in half. What used to need a small operations team — merchandiser, content person, junior dev for catalogue maintenance, plus periodic SEO consultancy — now runs with a much smaller core team plus the AI tooling. We kept the people who do the work that the tools can’t (strategic decisions, supplier relationships, customer experience design) and removed the work that didn’t need a person.Hundreds of hours saved per month. Conservative estimate, looking at the recurring workflows that the tooling now handles. Clearance triage was ~40 hours/week previously, run quarterly. Bulk catalogue operations used to be a multi-day project; now it’s a script. Content for new collections used to take 1-2 days each; now closer to 1-2 hours.Catalogue health. Every one of the 7,000+ variants has been touched in the last year — created, edited, repriced, retagged, metafield-tuned, content-rewritten. Most stores at this scale have stale tail-end inventory accumulating dust. We don’t, because the tooling makes it cheap to maintain.SEO foundation. 200+ collections with rewritten and optimised content. Hundreds of meta titles and descriptions reworked at scale. The content depth wouldn’t have been feasible for a team this size pre-AI. Our Shopify SEO service applies the same patterns to client stores.Operating cost. Running the AI tooling costs a fraction of what the headcount it replaces used to. Claude API spend across all the agents and skills is well under what one mid-level salary would cost.We’re not claiming this is the only way to run a Shopify Plus store. We are saying: it works, it scales, and the team running the store now is happier doing the strategic and creative work that’s left to them.

What it means for client engagements

The reason we put this case study on the site is straightforward: the AI engineering work we sell to clients was built here first. We use it daily. We know where it works, where it doesn’t, and how long each part took to get right.When we deploy a similar stack into a client’s Shopify Plus store, we’re not learning on the job. We’re transplanting workflows that already run a store we own.Most of our AI engineering engagements involve some subset of what’s described above — typically 8 to 12 weeks to ship the agents, MCP servers and skills that compound the most for that specific store. The deliverable looks different for every client because every catalogue, customer base and operating model is different. What’s consistent is the engineering pattern: in-house code, owned by you, deployed inside your environment, ready to run after we leave.If you run a Shopify Plus store doing £1m+ and you’ve read this far, the conversation is worth having.

See the AI engineering engagement model →Back to all case studiesTalk to us