What Is ROI? How to Calculate It for Design Projects and Present It to Your CFO

What is Return on Investment, exactly?

ROI is a simple ratio with a centuries-old pedigree: how much did this investment return, relative to what it cost?

ROI = (gain from investment − cost of investment) ÷ cost of investment × 100

A few properties matter when designers use it:

• It’s a ratio, not an amount. “SAR 300,000 of value” is a claim; “150% ROI with 8-month payback” is a comparison finance can weigh against every other project competing for the same budget — which is exactly how they’ll use it.

• “Gain” must be money, or convertible to money. Revenue gained, costs avoided, hours saved priced at loaded rates. Satisfaction scores and NPS don’t enter the formula; they support the story around it.

• ROI says nothing about time. That’s why it travels with a companion metric: the payback period — how many months until the cumulative return covers the cost. A 300% ROI over five years loses to a 100% ROI over four months in most boardrooms.

How do you calculate ROI for a design project, step by step?

Step 1 — Define the metric the project moves. One primary, two secondary at most. A checkout redesign moves conversion rate. An onboarding redesign moves activation. A support-flow redesign moves ticket volume. If you can’t name the metric, the project isn’t ready to be priced.

Step 2 — Baseline it. Current value, measured over 8–12 weeks. Conversion 1.8%, tickets 450/month, activation 34%. Whatever the number is, write it down with a date — this is the anchor every later claim hangs from.

Step 3 — Estimate the improvement, conservatively. Use the low end of whatever evidence you have: previous projects, published benchmarks, pilot results. A 5–15% relative lift is a defensible range for most conversion work; 20%+ claims need strong evidence attached. The conservative estimate isn’t modesty — it’s what makes the model believable.

Step 4 — Convert to money. Multiply through by volume and unit value:

Monthly value = metric improvement × volume × unit value

Example: 1.8% → 2.1% conversion on 25,000 monthly transactions at SAR 200 average order = 0.003 × 25,000 × 200 = SAR 15,000/month, or SAR 180,000/year. Add secondary lines if they’re solid — ticket reduction, rework avoided — each computed the same way (the testing ROI calculator article has all four formulas worked).

Step 5 — Apply the formula. Project cost SAR 120,000, annual value SAR 180,000, attribution 60% → attributed value SAR 108,000. Year-one ROI = (108,000 − 120,000) ÷ 120,000 = −10%… which is why you also show year two: (108,000 + 108,000 − 120,000) ÷ 120,000 = 80% at 24 months, with payback around month 14. Honest multi-year math beats an inflated year one.

How do you present the estimate to finance and the CFO?

This is where most design budget requests die — not on the math, but on the presentation. The rules that work:

Show three scenarios, not one number. Conservative / expected / optimistic, side by side. One-number estimates signal advocacy; three-scenario estimates signal analysis. CFOs fund analysis.

State your assumptions where they can be attacked. “We claim 60% attribution because a pricing change ships in the same quarter” — put it in the document. An assumption that’s visible can be negotiated; one that’s discovered destroys the whole request.

Lead with payback, support with ROI. “This pays back in 9 months” is the sentence that lands. The ROI percentage is the footnote that survives scrutiny afterward.

Price the downside. “If we’re wrong and the lift is half our conservative case, payback extends to 18 months and the project still breaks even by year two.” The request that names its own failure mode is the one that gets approved — because it’s the only one that sounds like it was written by someone who understands risk.

Speak cost-language, not design-language. Never say “delight,” “seamless,” or “brand elevation” in the finance deck. Every line is revenue, cost, or risk. (We wrote a whole piece on this translation: How to Explain UX to Your Board.)

Bring evidence, not analogies. One filmed session of a real user failing the current flow justifies more budget than any benchmark. Pair the spreadsheet with the clip: the numbers prove the size, the clip proves the cause.

What mistakes get design estimates rejected?

• Benchmarks as the case. “Forrester says UX returns 100:1” is context, not evidence. Your funnel data is the evidence.

• 100% attribution. Claiming the entire lift when three other things changed that quarter. It works once, then your credibility is gone for the next ask.

• Estimating without a baseline. “We think conversion will improve” against an unmeasured present is a guess in both directions.

• Hiding the cost of doing nothing. Actually — this one cuts the other way: teams forget to include it. “Maintaining the current flow costs SAR 86,000/year in tickets” is often the strongest line in the whole document. Use it.

If you want the estimate built properly — baselined metrics, priced problem list, scenarios that survive a finance review — that’s the analytical core of our UX strategy consulting. Book a strategy call →