Book My Diamond Suite
Sales Expert App + Admin App
Built from scratch to turn in-store diamond selling into a fast, confident, and educated experience, for salespeople and customers alike.
iOS + Android
Salesperson (primary), Admin/Owner/Purchase Head (secondary)
Product designer + PM + content (solo)
CEO (stakeholder)
2 engineers
Dense diamond data on phone, readability, offline states, Ionic framework
PostHog (events instrumented; tracking live)
PM Summary
Problem
In-store diamond selling relied on broker calls, manual orders, and undertrained staff. Customers waited 20 to 30 mins or left for another store.
Goal
Replace calls and guesswork with live inventory, guided discovery, and in-app knowledge so every salesperson sells like an expert.
Ownership
End-to-end product design, PM, and content, reporting to the CEO and working with 2 engineers.
Shipped
20+ features across Sales Expert and Admin apps, full design system built from scratch.
Impact
Pre-booking calls reduced ~80% (customer feedback, 8 months);
PostHog analytics live for deeper tracking.
What we saw
Traditional in-store diamond selling had three breaking points
Limited stock, long waits: Stores kept 3 to 5 physical stones. Anything specific meant a 20 to 30 min broker call and customers walked out before it ended.
8-year-old tool, not built for the counter: The legacy BMD app was a back-office trading tool. Showing it to a customer in-store felt unprofessional and slow.
Undertrained salespeople: Diamond education takes years. When customers asked technical questions, salespeople either guessed or called for help, both eroded trust.
What we changed
60–70 jewellers interviewed at B2B events + large store owner conversations. Three insights anchored every decision.
Customers walk in with 3 types of intent
Budget-first (gift buyer with a number in mind), 4C-aware (knows carat/cut/colour), spec-specific (knows exactly what they want).
We built 3 separate discovery paths instead of one search.
Salespeople felt stuck on technical questions
Every parameter question was an exit risk.
We built Knowledge Book: image-led explanations accessible from diamond parameters.
Cart gave order power to the wrong person
Wrong stones got ordered and admins had to verify on calls.
We replaced cart with Request-to-Book, making admin the purchase decision-maker.
Decide
Three decisions that shaped the product, each went through at least 2 rounds before locking.
Decision 1: Showroom design (iterated 3 rounds)
Round 1
What We Built
Basic diamond card with a label showing its category, e.g., engagement rings
Users Said
"Feels like one stone, not a collection"
Round 2
What We Built
Visual tray that looks like the tray used in the jewellery stores, one diamond per tray
Users Said
"Feels like one stone, not a collection"
Round 3
What We Built
Multiple diamonds per tray, each with carat + quality + price label (what users understand or want to know)
Users Said
"This looks like our actual stock tray"
Decision 2: Two search modes, not one
Smart Search
Problem
First-time users often don’t know how to use filters and end up selecting combinations that return no results.
What we built
Smart Search expands strict filters automatically to keep results flowing.
Why it worked
No dead ends at the counter; faster time-to-first-shortlist.
Intelligent Search
Problem
Exact-match searches don’t always show the best value; salespeople may not know the “industry tricks” to find better stones.
What we built
Intelligent Search suggests stronger options by slightly adjusting criteria (grade blend, size trade-offs, alternate grading, etc.).
Why it worked
Helps salespeople sound like experts and gives customers better options without extra calls.
Decision 3: Cart out, Request-to-Book in
Problem
Cart gave purchase power to the salesperson; admins had to call back to verify or fix wrong orders.
What we changed
Removed cart from Sales Expert.
Added Request‑to‑Book: sales creates a request with a note (“customer confirmed, wants this stone”), then admin reviews and books in Admin.
What we gained
Clear handoff: salesperson recommends, admin decides.
Fewer wrong stones ordered.
~80% fewer calls between sales and admin (reported by stores).
What we shipped
The final solution was split across two connected apps: a Sales Expert experience for in-store selling and an Admin experience for review, control, and order management. Together, they turned the product from a simple trading tool into a more complete sales workflow.
Sales Expert App
Showroom for curated discovery.
Redesigned search, filters, and results.
Smart Search and Intelligent Search.
Shortlist flow for customer-ready options.
Request-to-Book for handing off confirmed selections.
Knowledge Book integrated into the selling journey.
Admin App
Request review and booking flow.
Markup controls for better pricing flexibility.
Order tracking and invoice visibility.
Role-based access for different team members.
New checkout details like saved addresses and billing information.
System-level improvements
Built a reusable design system from scratch.
Created a modular structure so features could be enabled or hidden by role.
Improved the overall flow between sales, admin, and order management.
What we learned
Designing for assisted selling is different from standard ecommerce, because the interface has to support both the salesperson and the customer in the same moment.
Dense inventory experiences work better when filtering, discovery, and decision-making are separated into clear steps.
Building a modular system early made it easier to support both Sales and Admin workflows without the product feeling disconnected.
Conclusion
This project helped transform Book My Diamond from a complex inventory-led experience into a more guided sales workflow across Sales Expert and Admin. By simplifying discovery, improving handoff, and structuring the system for scale, I was able to ship a product that better supported real showroom conversations and internal operations.