User Research & Empathy

LIMS Transformation

How watching users turned a 15% adoption disaster into 95% success. Stakeholders blamed training. Users blamed the system. I found the real problem.

PM Skills Demonstrated
  • User Research
  • Process Mapping
  • Architecture Design
  • Stakeholder Management
LIMS Transformation

The User Problem

When Cotecna acquired a consumer product testing lab business in 2022, they inherited a LIMS (Laboratory Information Management System) that users had essentially rejected.

The numbers told the story: only 15% adoption. Lab technicians were using the system for one thing—registering products for testing—and then immediately abandoning it for their real workflow: paper pages, Excel spreadsheets, and Word documents.

Who were these users?

  • Registration staff who logged incoming products
  • Lab analysts who performed 2,000+ different tests across 4 testing standards
  • Report specialists who compiled results into client-facing documents

The lab analyst would conduct tests and write results on paper pages. The report person would physically collect these pages from around the lab, then manually type everything into a Word document—one value at a time. Save as PDF. Print for signatures. Repeat for every single report. A report that took 30 minutes to test took hours to document.

The Insight

Stakeholders had a diagnosis: "Lack of adoption"—a fancy way of saying "users need training."

But I wasn't convinced. If training was the answer, why were the same users successfully using Excel and Word? Those aren't simple tools.

I went hands-on. I sat with each team member at each level—registration, analysis, reporting—watching how they actually worked. Not asking what they wanted. Watching what they did.

What I found wasn't a training problem. It was an architecture problem.

Consumer product testing involves roughly 2,000 different tests. Each test can follow 4 different testing standards. That's effectively 5,000+ unique combinations, each with its own reporting tables, parameters, values, and calculations.

The users hadn't rejected the concept. They'd rejected the inflexibility. The paper-and-Excel workaround wasn't inefficiency—it was rational adaptation to a system that couldn't handle their reality.

The Decision

I had two paths:

Option A: Keep adding test configurations to the existing system. This is what the technical team proposed. It would take months, require constant maintenance, and still wouldn't handle edge cases.

Option B: Redesign the architecture to let users handle variation themselves.

I chose Option B, but it required convincing stakeholders to abandon two years of their previous approach.

What I Said NO To

  • Adding more hardcoded test configurations
  • Building a "training program" to force adoption
  • Incremental improvements to the existing workflow

What I Pushed For

  • Database restructure to support dynamic configurations
  • A form builder that lets users create their own reporting tables
  • Connecting form-built tables directly to tests in the system

Strategic Impact

15% → 95%

System Adoption

Users didn't need training. They needed a system that worked with them, not against them.

30 Days

Time to Transform

Rolled out to 4 labs within one year, managing 100% of operations.

Ankit's expertise in product development, stakeholder management, and team collaboration was instrumental in driving project success. His keen understanding of customer needs and ability to translate them into actionable product roadmaps was impressive.

The Lesson

How I apply this now:

When users reject a product, the instinctive response is "they need training" or "they'll adapt." Sometimes that's true. But often, user rejection is intelligent feedback about a fundamental mismatch between the system and their reality.

The diagnostic question I always ask: Are users successfully using other tools to accomplish the same goal? If yes, the product has a design problem, not a training problem.

My methodology: Don't ask users what they want. Watch what they do. The workarounds reveal the requirements.

The architecture principle: If a domain has high variation (thousands of test types, customer-specific requirements, edge cases), don't try to hardcode every variant. Build systems that let users handle variation themselves.