You know those tools everyone says are "important" until they actually try to use them?
That was NCRs in Visibuild.
In construction, a Non-Conformance Report (NCR) isn't a checkbox. It's a record. A safety net. Sometimes, legally speaking, a lifeline. But in practice, our NCR feature wasn't living up to that reality.
Before this project, NCRs in Visibuild technically existed, but barely functioned in a way customers could rely on. If you treated them like inspections and ignored most real-world complexity, they kind of worked. But that's not how quality is managed on real sites.
The truth was harder to ignore: the solution was so weak that customers actively avoided it.
After trying to use NCRs in Visibuild, teams either moved to other platforms that handled quality workflows properly, or fell back to paper forms and Excel spreadsheets, the tools they trusted, even if they hated them.
That's the worst signal a product can get. Not complaints. Not feature requests. Silence, abandonment, and workarounds.
The Problem
You know those tools everyone says are "important" until they actually try to use them?
That was NCRs in Visibuild.
In construction, a Non-Conformance Report (NCR) isn't a checkbox. It's a record. A safety net. Sometimes, legally speaking, a lifeline. But in practice, our NCR feature wasn't living up to that reality.
Before this project, NCRs in Visibuild technically existed, but barely functioned in a way customers could rely on. If you treated them like inspections and ignored most real-world complexity, they kind of worked. But that's not how quality is managed on real sites.
The truth was harder to ignore: the solution was so weak that customers actively avoided it.
After trying to use NCRs in Visibuild, teams either moved to other platforms that handled quality workflows properly, or fell back to paper forms and Excel spreadsheets, the tools they trusted, even if they hated them.
That's the worst signal a product can get. Not complaints. Not feature requests. Silence, abandonment, and workarounds.
My Role
I was the product designer on this project, working across workflow design, interaction patterns, table behaviour, and information hierarchy. I collaborated closely with PMs and engineers to ensure what we designed wasn't just conceptually "right," but usable under real construction pressure.
Research: Learning How NCRs Actually Work
We didn't guess. We listened.
We ran interviews and workflow walkthroughs with teams at Icon Group, Texco Construction, Southbase, and others. We reviewed real NCR registers. Real spreadsheets. Real workarounds.
What emerged was a strikingly consistent picture across very different organisations.
What Customers Told Us
On the current system:
"Current system just allows requirements dump, very not good." — Mina & Josh, Icon Group
"Current implementation lacks proper workflow. Doesn't flow like standard checklists. Missing multi-reviewer stage process." — Jonathan Glick, Texco Construction
On what they actually needed:
"Register view is 'way better' for having all information accessible. Critical for calling subcontractors about incomplete NCRs." — Ezgi, Icon Group
"Register view ideal for: Senior management reporting, subcontractor meetings, monthly client reports." — Esther, Icon Group
On scale and severity:
"NCRs are 'much bigger' than inspections/observations. Class 1 NCRs require investigation similar to safety incidents." — Patrick Woodbourne, Southbase Construction
"Used for critical level issues, not routine defects. Only raised for serious failures: steel, concrete slab, earthwork." — Jonathan Glick, Texco Construction
Cross-Customer Patterns
Three problems appeared consistently across every customer we spoke to:
Template structure was broken
Multiple customers described templates as "unwieldy" when everything sat under requirements. They needed hierarchy: the ability to group tasks within NCRs.
Workflows didn't match reality
NCRs involve multiple companies, formal approvals, and clear accountability. They don't close themselves. They don't move in straight lines. And accidentally closing one is a serious problem. Our system had none of this.
Registers mattered more than forms
Managers don't live inside individual NCRs. They live in registers, scanning status, ownership, and risk. Customers were manually rebuilding NCR registers in Excel because nothing in Visibuild could replace them.
Mapping the Experience
To understand where the existing system broke down, I mapped the end-to-end experience for two core workflows: raising NCRs after site incidents and tracking NCRs across a project.
Exploring Solutions
With the research synthesis complete, I explored several approaches to the tracker problem: the core interface where managers would spend most of their time.
The question wasn't just "what fields do we show?" It was: "What decisions are people trying to make every day, and what information do they need to make them?"
Option 1: Template-style Tracker (Progress View)
What it's good for:
- Managing the lifecycle of NCRs at a project level
- Teams can instantly see where each NCR is stuck
- Leans into our existing template/task model
Challenges:
- Harder to do structured reporting at project or company level
- Risk of becoming a glorified checklist rather than a learning tool
Option 2: Register-style Tracker (Reveal Answers)
What it's good for:
- Functions like a classic NCR register, anyone can scan through and see what went wrong and what was done about it
- Makes thematic reporting possible at the company level (e.g. "Top 5 recurring root causes across all projects")
- Strong alignment with existing paper/NCR workflows
Challenges:
- Can feel rigid if users want to adjust the flow
- Becomes heavy if NCRs have lots of steps or parallel requirements
Option 3: Hybrid Tracker (Toggleable View)
What it looks like:
Each NCR tracker, has a control to toggle the display mode. Requirements view shows the underlying text/answers captured in the NCR. Progress view flips the same row into a progress-style tracker.
What it's good for:
- Combines the clarity of a register with the visibility of a progress board
- Lets users zoom in on progress when managing day-to-day, then toggle to substance when doing reporting or analysis
- Reduces double entry: requirements and progress stay connected to the same underlying NCR template
Why NCRs differ from normal templates:
- NCRs appear sparingly on the Y-axis (sometimes 5 max per project, as noted by Texco), unlike templates that are duplicated in quantity
- Impact and friction are understandably high, no-one wants to see where they were displayed incorrectly
- The tracker needs to handle sparse, irregular NCR review while maintaining visibility across projects
What We Shipped
Based on the research and exploration, we landed on a hybrid approach with two distinct views: a Tracker view for progress-based lifecycle management, and a Register view for detailed, Excel-style reporting.
Key Design Decisions
Strengthen the workflow
Multi-step approvals. Manual close-out (no silent auto-closing). Clear ownership and status at every stage.
Standardise creation
Previously, NCRs varied wildly by project and region. We introduced a consistent structure that teams could rely on. Less flexibility, more trust.
Build a register-style tracker
Explicitly mirrors Excel's mental model. Rows = NCRs, columns = critical fields. Filtering by location, company, status. CSV export for deeper reporting.
We didn't fight Excel. We respected why users trusted it.
Impact
Enterprise adoption: Icon Group committed to company-wide NCR usage starting January 1st, replacing their existing system with Visibuild.
Less Excel: Teams stopped manually rebuilding NCR registers, the Register view gave them what they needed natively.
Better oversight: Management could finally see what was open, what was stuck, and why, without exporting data to another tool.
No vanity metrics. Just behaviour change.
Lessons Learned
Industry knowledge isn't optional
If you don't understand how work actually happens, users will abandon your product. NCRs aren't inspections. They require multi-party accountability, deliberate close-out, and audit trails. Missing any of that made the feature unusable.
Standardisation unlocks scale
Especially in enterprise tools, consistency often beats flexibility. When Icon mandated "no ad-hoc NCR creation," it wasn't bureaucracy, it was the only way to make quarterly reporting meaningful.
Mental models matter
Users trusted Excel for a reason. Respecting that, building a register that looked and worked like their spreadsheets, was the fastest path to trust. We didn't try to be clever. We tried to be useful.
Closing Thought
This project wasn't about adding features.
It was about fixing something so broken that customers avoided it entirely.
Those are the hardest, and most satisfying, problems to solve.