Context
NEST is a real estate consultancy operating in Vietnam's premium property market. When I joined, the sales team — a group of experienced property consultants — was managing their entire pipeline in WhatsApp threads and spreadsheets. Leads were lost between handoffs, follow-ups missed, and there was no visibility for management into what was happening in the funnel.
My role was unusual: I was brought in as Lead Product Designer but I also worked as a sales consultant, using the spreadsheet system myself before building the replacement. This gave me direct experience of the domain problem.
Problem
The core challenge wasn't technical — it was workflow. Every consultant had their own system, their own shorthand, and their own informal process. A CRM that didn't map precisely to how consultants actually worked would be ignored regardless of how well designed it was.
The risk with a 0-to-1 product is building the wrong thing efficiently. I needed to understand the work before I could design the tool.
Process
Embedded discovery
For the first 6 weeks I shadowed consultants through their day — sitting in client meetings, watching how leads were captured on phones, how follow-up was tracked, how handoffs between junior and senior consultants were managed. I took no Figma notes. I took field notes.
The key insight: consultants didn't think in terms of pipeline stages. They thought in terms of client readiness — a qualitative judgement about how close a client was to committing. Any CRM that forced them into rigid funnel stages would feel like extra admin, not a tool.
Design system first
Before designing any screen, I built the design system. This sounds backwards but it was deliberate — the system forced me to make decisions about visual language, density, and component behaviour before I was attached to any specific layout. When I started building screens, I was composing from a library rather than inventing from scratch, which meant I could prototype faster and more consistently.
Iterative build with consultants
Every two weeks I showed working Figma prototypes to 3–4 consultants. Not wireframes — interactive prototypes that felt close to real. This pace surfaced issues early and kept the team bought in. Several consultants became informal champions for the tool before it launched.
The client readiness model
I designed a custom status model based on the field research. Instead of generic pipeline stages, the CRM used a 5-point "readiness" indicator that matched how consultants already spoke about clients. Status changes were one tap. The system was opinionated about what mattered — and flexible about everything else.
Outcome
The CRM launched and was adopted across the sales team. Lead tracking improved, follow-up rates increased, and management had their first real-time view of the pipeline. In the same period I was working as a sales consultant using the tool — I won Sales Employee of the Month in August 2024, partly by using the CRM to maintain follow-up discipline that competitors couldn't match manually.
The design system built for NEST is still in active use and has been extended to support new product areas.
Learnings
Don't design until you understand the work. Six weeks of field research felt slow. It saved months of redesign. The "client readiness" model came directly from listening to consultants talk, not from any design framework.
Build the system before the screens. A design system built upfront isn't overhead — it's speed. Every screen I designed came together faster because I wasn't solving component problems at the same time as layout problems.
Using your own product changes your design decisions. Working as a consultant and using the CRM daily surfaced issues that no amount of user testing would have caught. The best form of research for a workflow tool is being a real user.