There is a persistent stereotype: design is about beauty. Buttons, colors, fonts. Something you do at the end, when "everything else is ready."

In practice, every interface decision is a hypothesis with business consequences. Where to place a button, how many steps in a registration form, what an empty state looks like — all of this directly or indirectly affects conversion, retention, and support costs. It's just that this connection is not always tracked.

When people talk about business metrics, they usually think of marketing or product. But design is embedded in each of them.

Conversion. A form with seven fields converts worse than one with three — even if it collects the same information. This is not theory: A/B tests confirm this time and again. Visual emphasis, step sequence, tooltip wording — all of this changes the percentage of users who reach the target action.

Time-to-value. Especially critical for B2B. If a new user opens a product and doesn't know where to start — they will leave before seeing the value. Good onboarding shortens this path. Bad onboarding extends it to the point where the user gives up.

Support load. Every confusing interface creates support tickets. In one of the B2B projects we supported at ROOT CODE, 40% of support requests were related to the same section. Redesigning that screen — rephrasing labels, restructuring the hierarchy of actions — reduced the flow of requests by about half within two months after release. This is a direct saving on operational expenses.

Retention. A user returns to a product not because they "decided to," but because it's comfortable — they understand where things are and don't spend cognitive effort on orientation. Design is the infrastructure of this comfort.

The problem is not that companies ignore the impact of design. More often, the problem is that this connection is not tracked systematically.

Teams release features but don't look at what changed in user behavior afterward. Or they look, but only at activation — not at retention a week later. Or they don't compare anything at all because there is no basic analytics.

Design without measurement is working blind. Good design decisions are confirmed by data — and bad ones are too. But only if the data is collected.

The minimum stack that allows this: product analytics with user scenario tracking, regular usability testing sessions (even 3–5 users per month provide surprisingly many insights), and the habit of looking at metrics before and after every significant interface change.

One of the clearest examples is onboarding redesign. Companies often underestimate how much a "failed" first user visit costs. Acquiring one user in B2B is expensive. If they leave without figuring things out — the marketing budget is wasted.

If redesigning onboarding increases activation fr om 30% to 45% — that's not "UX improvement." That's 50% more users who reached the point wh ere the product begins to appeal to them.

The same applies to redesigns of analytical and administrative interfaces. When a user spends not 12 minutes on a task but 4 — they accomplish more during the workday. On the scale of a team of 50 people, that's several full working days per week.

- Define 2–3 key metrics you want to influence — before starting the design, not after;

- Set up at least basic tracking of user scenarios in your analytics;

- After each release, give interface changes 4–6 weeks and look at the numbers;

- Conduct at least one usability session per month — even informal, with colleagues from another department;

- Don't confuse "we feel it got better" with "we have data that it got better."

Design begins to impact business the moment a team stops perceiving it as decoration and starts perceiving it as a hypothesis. Every screen, every scenario, every button is an assumption about how the user will act. The designer's task is to make that assumption conscious and verifiable.