Design ROI: the 6-step calculation with a real project

Every founder has heard that design pays off. Few have seen the math. When money gets tight, design becomes the first line cut from the budget, precisely because no one can show what it returned.

On September 11, 2026, Alex Williams published a method in Smashing Magazine for building a design ROI case that holds up in a board meeting. The opening line is direct: “Strong UX ideas do not secure investment on their own.” Here I summarize the method and try to redo the math with one of our projects.

The article's example is made up, and that matters

The case Smashing uses to demonstrate the method is a B2B SaaS company called Meridian. The article itself notes it's entirely fictional. The numbers show the mechanics, not a market benchmark.

Still, the shape of the math is worth seeing. In the example, an onboarding redesign lifts trial-to-paid conversion from 8.0% to 9.4%, which adds about 560 paying customers a year. At an average annual revenue of US$1,800 per account, that's close to US$1 million in new revenue. The article attributes 70% of that gain to the redesign, as a stated assumption, and divides by the total investment of US$117k. The result is an ROI of roughly 5 to 1 in the first year.

The point isn't the 5:1. It's that every number has a source, and every assumption is written down.

The 6 steps of the method

The article doesn't number the steps, but the structure follows this order:

1. Define success. Talk to decision-makers in each area, find the recurring themes and write each one as a goal with a number. Favor metrics that already show up on the company's dashboards.
2. Add up the total cost. Not just design. Include engineering, tools, coordination and people's time in meetings and reviews, at their real hourly cost.
3. Prove causation. A/B test when you can. When you can't, use a time series with a baseline before the change. Log what else changed in the same period and decide the attributed percentage before seeing the result.
4. Calculate. Incremental gain times average revenue, times attribution, divided by investment. Secondary savings, like fewer support tickets, go on a separate line.
5. Tailor it to each decision-maker. Finance wants cost and risk. Marketing wants conversion. Product wants support volume. The numbers don't change between versions; only the order does.
6. Gather qualitative evidence. NPS, satisfaction, user quotes, collected systematically. And people inside the company who can retell the story.

Williams sums up the problem in one line: “Executives don’t hate UX, they just hate vagueness.”

Redoing the math on a real project: AI Vou Eu

Here's our case. I only use what's public about the project. Anything we haven't measured yet or can't publish stays out.

AI Vou Eu is an AI travel SaaS. With the Studio, the product went live in 4 weeks.

Going through the six steps:

1. Success. For a founder who hasn't launched yet, success isn't conversion. It's having the product live, with real users, on a known date. The business metric that comes after is the one the client defines: signups, bookings or revenue.

2. Total cost. On one side, the cost of going without a product, month after month. On the other, the project's fixed price, plus the client team's hours in meetings and reviews.

3. Causation. Here attribution is simpler than in the Smashing example. Before, there was no product. After 4 weeks, there was. Nothing else competes to explain the product being live. What takes care is attributing what came after, because other things change in the same period.

4. Calculation. The honest math has two parts. First, the avoided cost: every month without a live product was another month of spending with no return. Second, the gain: the result of the first period live divided by the total project investment.

5. Decision-makers. For a founder, the finance question is “how much more will I spend before I see something working?” With a fixed price after discovery and a deadline written into the contract, that answer exists before you start.

6. Qualitative. What the client and first users say about the product counts too, even without a number.

Why pre-launch ROI is different

Smashing's method was designed for companies that already have a product, traffic and a baseline. The redesign improves something that exists. For a founder with no live product, the first line of the math is different: what each week without launching costs.

Every dollar spent before launch earns nothing until the product ships. That's the easiest return to measure and the most ignored, and it's on the list of ways to burn money before launch. Before debating how many conversion points good design brings, the question is whether there's anything to convert.

What this means for you

If you're about to hire design or development, do the math before you sign:

• Write down the success metric. Just one, with a number and a date.
• Add up the total cost. Include your time and your team's, plus the price of a custom web system.
• Record the baseline today. Even if it's zero.
• State the attribution. How much of the result you'll credit to the project, and why.
• Count the cost of waiting. How much cash goes out each month without a live product.

Williams closes with an idea that applies to studios and clients alike: “repeatable ROI is what earns recurring investment.”

How we do it at the Studio

We start with discovery. After it come the fixed price and the deadline written into the contract, which settles steps 2 and 5 before day one. Every system ships with an AI agent that runs daily, flags failures and checks that the data makes sense, which helps keep the baseline alive after launch. The success metric gets written down with the client during discovery.

If you have a SaaS to launch and want the math in writing before you start, our SaaS and custom system development service starts with that discovery.