10,000+ customers
Enquiry volume that cannot be handled by manually reading and answering every message.
Case study
GPUsed: running a 10,000 customer business on automation.
GPUsed is the UK technology resale business Max Brocklesby runs. It is not a client. It is where these systems were built first, under real commercial pressure, before any of them were offered to anyone else.
The business
Real volume, owner-led.
GPUsed buys and sells used computer components across the UK. It has served 10,000+ customers and collected 1,000+ customer reviews.
1,000+ reviews
Reputation built from completed orders rather than paid promotion.
No large admin team
The constraint that forced the automation. Growth meant the owner became the bottleneck.
The bottleneck
The owner became the constraint.
The pattern is familiar to any owner-led business that grows. Enquiries arrive faster than they can be read. Quotes need answering while the work that earns money waits. Follow-ups get forgotten, so warm enquiries go cold for no reason other than time. Stock, admin and reporting fill the evenings.
None of that is a strategy problem. It is a throughput problem, and it is exactly the kind of problem automation solves without adding headcount.
The systems
What actually runs.
Each of these maps directly to a service Brocklesby builds for clients. They were built here first.
Enquiry handling
Inbound messages are read, classified and drafted against rather than queuing untouched. Inbox management.
Quoting
Pricing responses go out quickly and consistently instead of waiting for a free evening. Quote handling.
Follow-ups
Enquiries that go quiet get chased automatically, so revenue does not leak through forgetfulness. Automated follow-ups.
Review requests
Completed orders trigger review requests, central to how the 1,000+ review count was built. Review requests.
Stock and admin
Repetitive internal workflows run without manual rekeying between systems. Spreadsheet automation.
Reporting
Numbers are assembled automatically rather than rebuilt by hand each time they are needed. Process automation.
Human approval
AI drafts. People still decide.
Nothing here is a fully autonomous system that emails customers unsupervised. Sensitive, high-value and judgement-heavy steps keep a person in the loop. That is the same design principle applied to client systems, for the same reason: an automation that embarrasses you once costs more than the time it saved.
See human approval in AI automation for how that oversight is structured.
What this does and does not prove
An honest reading.
What it proves: these systems work at real volume in a real UK business, they were built by the person who would build yours, and the problems they solve are ones Max has had himself rather than read about.
What it does not prove: that your business will see the same result. GPUsed is a resale business with its own volumes, tools and processes. Figures shown across this site are typical, illustrative examples, not guarantees. Yours will differ. That is precisely why the first system is built and tested before you are billed for anything.
Free automation audit
Find your version of this.
Start a free audit and get a practical recommendation for the first system worth building.