Internal tools lose funding arguments for a structural reason. A customer-facing feature arrives with a revenue projection. A workflow fix arrives with a complaint. Complaints do not compete well with projections, so the workflow stays broken for another year.
The fix is not better advocacy. It is arithmetic.
Watch the work before you model it
Do not start with a process diagram from a wiki. Start by sitting with four or five people while they do the task, and record what actually happens.
The gap is always large. Documented processes omit the spreadsheet that holds everything together, the colleague who gets asked because a screen is ambiguous, the copy-paste between two systems that were supposed to be integrated in 2022, and the second monitor holding a reference table.
Nobody documents the workaround. The workaround is where the money goes.
Count three things
Most workflow cost reduces to three measurable quantities.
- Time on task: minutes per instance, times instances per week, times people doing it.
- Rework: how often the task is redone, and what each redo costs.
- Delay: how long the item waits between steps, and what waiting costs downstream.
Time is the easiest and the least persuasive on its own. Rework is where the surprising numbers live, because errors introduced early are caught late by someone senior. Delay matters most when the queue sits in front of revenue, such as onboarding, credit approval, or provisioning.
Convert carefully, and stay conservative
Turning hours into money invites scepticism, so make the assumptions boring and explicit. Use fully loaded cost. Do not claim that saved minutes become billable output automatically. State clearly which savings are capacity released rather than cash saved.
A defensible smaller number beats an ambitious one. The moment a CFO finds one inflated assumption, the whole case is discounted. We deliberately present the conservative figure and note the upside separately.
A worked example
On a healthcare engagement, coordinators moved between six systems for one patient handoff. Observed time was 34 minutes, of which about 22 were switching, re-entering data already present elsewhere, and confirming details by phone because two systems disagreed.
Eleven coordinators, roughly nine handoffs each per day. The switching cost alone was close to two full-time roles, before counting the errors that caused rework further along the pathway. Nobody had ever added it up. Once someone did, the redesign was funded in a fortnight.
Include the cost of change
A credible business case also carries the other side of the ledger: build effort, migration, training, and the temporary dip while people learn the new flow. Leaving that out is the fastest way to lose trust with the person approving the budget.
It also forces an honest conversation about adoption, which is where most workflow projects actually fail. A better process that half the team ignores is not a saving. It is a second process.
This is the thinking behind our workflow and operations consulting work. Written by the UXperts team. If this is a problem you are living with, tell us about it.