Fifteen years inside PNC, Bank of America, Deloitte, and KPMG taught me something uncomfortable about design: being right isn't enough, and being loved isn't either. Budget goes to what gets measured. And design teams — mine included, early on — keep walking into budget meetings armed with adjectives.
The numbers exist. Designers just don't claim them.
Design work moves real metrics — I've spent a career watching it happen. A 40% engagement boost and 25% satisfaction lift on a banking platform. A 15% error reduction. A 30% satisfaction jump in healthcare. Not one of those numbers announced itself as 'design impact.' Somebody had to draw the line from the design decision to the metric — and if designers don't draw it, nobody does. The work gets filed as a cost line, and next year's budget gets smaller.
Good work loses budget fights to weaker work with better spreadsheets.
Model the levers, not the pixels
Design's business case lives in a handful of levers — conversion, retention, support load — and you can model them honestly with an editable projection instead of a promise. Two rules, straight from watching real budget meetings. One: a credible first estimate has to take 30 seconds, because anything that needs an afternoon never gets opened. Two: the output has to be board-ready, because a number only matters if it can travel without you in the room. This annoyed me enough that I eventually built a little calculator for it — but honestly, a napkin works. The discipline is the point, not the tool.
Build the number from what the problem already costs
Most design business cases fail because they start from the upside, which nobody believes. Start from the cost that already exists, which nobody disputes. Support tickets about one confusing screen — someone can pull that count. Hours a team spends re-keying data because two systems don’t talk. The percentage of people who start a flow and never finish it. Those numbers are usually sitting in a tool your company already pays for, and nobody has bothered to put them in the same sentence as a design proposal. Once you have the cost of the status quo, your ask stops being “design is valuable” and becomes “this is what it costs to keep the current version.” One of those is an opinion. The other is arithmetic.
Give a range, and show the assumptions
A single confident number invites an argument about the number. A range with visible assumptions invites a conversation about the assumptions, which is the conversation you want. Say plainly: if we cut this flow from six steps to three, and completion improves somewhere between two and eight percent, here’s what that’s worth at current volume — and here’s the volume figure I used. Now a finance person can push on your inputs instead of dismissing your conclusion. In fifteen years inside banks I never once saw a CFO reject an estimate for being uncertain. I saw plenty rejected for having no visible math behind them. Uncertainty is normal. Unshowable reasoning isn’t.
What to do in your next budget cycle
Before the work: write down which lever the design is supposed to move, and what it's worth if it moves. After the work: measure it, and say the number out loud, in the room where budgets actually get decided. That's the whole discipline. Design didn't earn its seat by being craft. It earns it by being the highest-leverage line item nobody was bothering to measure.