A jewelry agency. 24 stores. Zero automation.
The client was a marketing agency specialising in D2C jewelry brands across the United States. They were not managing one store — they were managing 24 jewelry stores simultaneously, handling everything from finding new leads to booking in-store appointments.
At the time we started working together, their entire operation ran on manual effort. A human was responsible for following up every single lead, booking every appointment, sending every reminder, collecting every review. Across 24 stores, this was not just inefficient — it was breaking the business.
The agency needed a system that could handle the full customer journey for every store without adding headcount. One system. Replicable across all 24 locations. Built in two weeks.
Six things that were quietly killing the agency.
What I built — step by step.
The entire system was built inside GoHighLevel over 14 days. The goal was not just to solve the current problems — it was to build a system replicable across all 24 stores with minimal setup per location.
How the full automation flow works.
What happened in the first 45 days.
The system went live across all 24 stores within 14 days of starting the build. Results started coming in within the first week.
Why this worked when others had failed.
The agency had tried to fix their lead follow-up problem before. They had hired virtual assistants. They had set up basic email sequences. Neither worked at the scale they needed.
The difference in this build was building for replication from day one. Instead of solving the problem for one store and then copying it 23 times, the entire system was architected as a single master setup that could be deployed to any store by changing a handful of variables — store name, location, calendar, phone number.
The second key decision was using AI calling instead of SMS-only follow-up. For jewelry purchases — which are high consideration, high emotion decisions — a voice conversation converts significantly better than a text message. The AI agent could answer questions about gemstones, pricing, and customisation in real time, which moved leads from interested to booked in a single call.
The third element was the no-show recovery workflow. Most automation systems treat a no-show as a dead lead. This system treated it as a different stage — with its own sequence, its own tone, and its own pathway back to a rebooked appointment. That single workflow recovered appointments that would otherwise have been written off.