Legacy software doesn't need a total reinvention
It needs taste, restraint, and a better visual system. The hard part isn't making it look new — it's making it feel current without breaking the habits people rely on.
A lot of legacy software does not need a total reinvention. It needs taste, restraint, and a better visual system.
The hard part is not making it look new. The hard part is making it feel current without breaking the habits people rely on.
This is one of the most common briefs I see in the wild. Some product has been alive for ten years. It serves a real audience that knows where everything is. Internally, the team is embarrassed by it. The interface looks dated. The buttons are inconsistent. The type scale was set up by someone who left in 2019. And every time the product team brings it up, the conversation defaults to "we should redesign it" — which means rebuild it, which means a year of work, which means it never happens.
The trap is treating a systems problem as a screen problem.
Almost every "this looks old" complaint is actually a systems complaint stacked across years of one-off decisions. Type choices that don't compose. Spacing values that drift between sections. Random icon styles. Six different button treatments. A color palette that was extended without a plan. Each individual thing is fine. The accumulation is what reads as "outdated."
The fix is rarely to redraw the screens. The fix is to lay down a real visual system underneath what's already there — clear type scale, consistent spacing tokens, a rationalized component library, an honest set of color decisions — and then let that system propagate through. Most of the screens don't need redesigning at all. They need re-rendering against the new tokens.
This is unsexy work. There's no big "before/after" reveal because the after looks suspiciously like the before. Same screens. Same flows. Same words. But everything fits now. Type sizes relate. Spacing is rhythmic. The buttons all came from the same place. The whole product feels twenty IQ points sharper without anyone consciously noticing why.
The productive question isn't "what does this need to look like in 2026?" It's "what's the smallest visual system that would let everything in here belong to the same product?"
That question is much cheaper to answer than a redesign, and the answer is usually most of the work.