Case Study
From nine spreadsheets to one source of truth
U.S. healthcare provider — CRM platform · Associate Project Manager · 2023
At a glance
Role
Associate Project Manager (Level I)
Duration
~6 months (2 months discovery + 4-month delivery plan)
Team
11+
Stack
.NET / Azure
Origin
Existing in-house MVP, no client request
The situation
Nobody asked for this project. We had an in-house MVP sitting unused, and I recognized that it addressed a problem I'd watched the client's Sales and Account Management teams live with: every person tracked their pipeline in their own spreadsheet. There was no shared process and no single version of the truth about any deal.
The challenge
Two things made this harder than a standard build. First, there was no existing process to digitize — the workflow between Sales and Account Management had to be designed from scratch through discovery, not documented from observation. Second, I had to sell the concept before I could deliver it. This wasn't a client saying "build us a CRM"; it was me arguing that they needed one.
What I did
Spent two months on requirements, deliberately.
On a six-month project, giving a third of the timeline to elicitation is an unusual allocation and I defended it. Building the wrong workflow fast would have been worse than building the right one slowly — and with no existing process to copy, the discovery was the design work.
Designed the funnel and the reporting model.
Defined the Sales to Account Management handoff and the single-source-of-truth reporting layer that replaced the spreadsheets.
Built a prototype clients could break.
I partnered directly with Design on a client-facing Figma prototype at high enough fidelity that stakeholders reported "bugs" in dropdown behaviour — inside the mockup. That misunderstanding was the best possible outcome: it meant we were collecting real UX feedback weeks before a line of production code was written.
Held a 4-month delivery plan with milestone tracking.
Tracked via Microsoft Project Gantt charts and roadmaps.
The outcome
- Every milestone hit on schedule across the 4-month delivery plan.
- Clean production go-live with an 11+ person team.
- Fragmented per-person spreadsheet tracking replaced with a centralized system.
- Prototype fidelity high enough that stakeholders engaged with mockups as the live product — UX feedback surfaced weeks pre-development.
What I took from it
The strongest projects I've run started with noticing a problem nobody had escalated. The spreadsheet situation was normal to everyone inside it; it took an outside view to name it as a problem worth six months.