Three questions to ask before you redesign anything

Redesigns fail in predictable ways. Most failures are not from craft — the work looks good — they are from chasing the wrong problem. Three questions, asked honestly at the start, catch 80% of those failures.

1. What metric, specifically, are we trying to move?

If the answer is “engagement” or “delight,” stop. Those words let everyone agree without agreeing. Force specificity: completion rate of the signup flow, seven-day retention, tickets in the “I cannot find X” support category. If you cannot name the metric, you cannot tell whether the redesign worked.

2. What is the cost of doing nothing?

Sometimes the answer is: not much. The current design is fine, it is not the bottleneck, and the team is bored. That is real, and the answer might be “do something else.” Quantify the cost of the status quo before quantifying the cost of change.

3. Who loses if this ships?

Every redesign has losers — internal teams whose workflow changes, external users who relied on the old patterns, dependencies that break. List them before you design. The redesigns that ship clean are the ones that addressed the losers explicitly, not the ones that pretended the losers did not exist.

If you cannot answer all three

You are not ready to redesign yet. That is fine. The hour you spend answering these questions saves a quarter of work.