I've watched a procurement officer spend 14 clicks to do something that should take three. I've watched a team lead keep a personal spreadsheet because the "official system" couldn't show her what she needed. I've watched an entire department quietly stop using a tool that cost six figures to build. Bad enterprise UX isn't a design problem. It's a business problem with a dollar figure attached.
What You Need to Know
- Every hour of bad UX multiplied by every user multiplied by every working day is a real cost that shows up in payroll
- Training costs for poorly designed systems can exceed the original development budget within two years
- Shadow IT, where staff build workarounds in spreadsheets and email, is the clearest signal that the UX has failed
- Good enterprise UX doesn't mean consumer-grade polish. It means the system works the way the person thinks
The Maths Nobody Does
Enterprise software decisions are made on features and price. Can it do X? Does it integrate with Y? What's the licence fee? The one question that rarely gets asked is: what will this cost us in lost time?
$100
return for every $1 invested in UX design
Source: Forrester Research, The Six Steps for Justifying Better UX, 2016
Forrester's research puts the return on UX investment at roughly 100:1. That number sounds too clean to be real, and it probably is in most contexts. But the direction is right, and the reason is simple. Enterprise software is used for hours every day by people whose time is expensive. Small inefficiencies compound fast.
Take a system used by 200 staff. If bad UX costs each person 15 minutes a day in unnecessary clicks, searching for information, and working around limitations, that's 50 hours of lost productivity per day. At an average loaded cost of $50 per hour, that's $2,500 per day. Over a year, $650,000. For 15 minutes.
Most bad enterprise UX costs more than 15 minutes. It's just invisible because nobody's measuring it.
The Three Costs
Training cost
Well-designed software teaches people how to use it. Poorly designed software requires training programmes. I've seen organisations budget $50,000 for training staff on a system that cost $200,000 to build. Then another $30,000 when the next cohort joins. Then again for the update.
3x
longer onboarding time for systems with poor usability vs well-designed alternatives
Source: Nielsen Norman Group, Enterprise UX Research, 2016
Nielsen Norman Group's research on enterprise usability shows that poorly designed systems take significantly longer to learn. That's not because the users are slow. It's because the system doesn't match how they think about their work. They have to learn the system's logic instead of the system matching theirs.
Good UX isn't free. But the cost of designing well is paid once. The cost of designing badly is paid on every new hire, every process change, and every annual refresher.
Adoption cost
The hardest cost to measure is the system that gets built and then not used. Not formally abandoned, just quietly ignored. People find ways around it. They keep their old spreadsheets. They email attachments instead of uploading to the portal. They call instead of logging a ticket.
This isn't resistance to change. It's rational behaviour. If the old way is faster and less frustrating, people will use the old way. They'll do the minimum required in the official system to keep management happy, and do their actual work elsewhere.
I've worked on projects where the client specifically asked us to "fix adoption." When we looked at the system, the adoption problem was obvious. The interface required knowledge the users didn't have. The workflow assumed a process that didn't match reality. The error messages were written by developers for developers. The system was hostile to the people it was built for.
Shadow IT cost
Shadow IT is what happens when adoption fails. Staff build their own tools. Shared drives full of Excel workbooks. Google Sheets tracking what the system should track. Personal databases in Access. Email threads serving as approval workflows.
Shadow IT isn't just inefficient. It's a governance risk. Data lives outside the official system, which means it's outside backup policies, access controls, and audit trails. When someone leaves, their spreadsheet leaves with them.
Shadow IT is the honest feedback that enterprise software doesn't want to hear. Full stop.
Rainui Teihotua
Chief Creative Officer
What Good Enterprise UX Looks Like
Good enterprise UX isn't about making things beautiful. It's about making things match the mental model of the person using them.
Match the workflow, don't dictate it. The system should follow the way people actually work, not the way a business analyst thinks they should work. This means spending time with actual users before designing anything. Not in a meeting room. At their desk, watching them work.
Reduce cognitive load. Show the right information at the right time. Don't show everything on every screen. Don't require people to remember things from three screens ago. Don't make them learn a navigation system more complex than the work they're doing.
Design for the 80% case. Most enterprise interactions follow predictable patterns. Design the interface for the common path, and make the edge cases accessible without cluttering the main view. The person who does this task 50 times a day shouldn't have to wade through options they never use.
Give feedback. When someone does something, the system should respond immediately and clearly. Did it work? What changed? What happens next? Silent submission forms and spinners that offer no indication of progress erode trust in the system.
The Investment Case
Enterprise UX isn't a luxury line item. It's the difference between a system that delivers its intended value and one that becomes expensive furniture.
The conversation with budget holders isn't "can we make it prettier?" It's "would you rather spend $40K now on user research and design, or $650K per year on lost productivity?" Those are real numbers from real projects.
When Isaac and I scope enterprise projects at IDESIGN, the user research and design phase isn't optional. It's not a nice-to-have that gets cut when the budget is tight. It's the phase that determines whether the other $200K gets wasted or not.
The technical build will work. We'll make sure of that. The question is whether anyone will want to use it. That's the question UX answers, and it needs to be answered before the build starts.