A 50% return beats a 20% return — until you notice one took five years and the other took one. Annualized, that 50% is worth about 8.4% a year. The 20% is worth 20%. The bigger headline number is the worse investment, and the plain ROI formula will never tell you that.
The formula, and where people slip
ROI — return on investment — is genuinely simple arithmetic:
ROI = (gain − cost) / cost × 100
"Cost" is what you put in. "Gain" is the profit: what you got back minus what you put in. Spend $10,000 on a project, get $13,000 back, and your gain is $3,000:
(13,000 − 10,000) / 10,000 × 100 = 30%
The slip is hiding in that word "gain." Plenty of people plug in the full $13,000 they received instead of the $3,000 profit and report a 130% ROI on a project that actually returned 30%. Both readings of "gain" are defensible in plain English, which is exactly why the number gets abused in pitch decks. Protect yourself by always writing the long form: (money out − money in) / money in. If "money out" is the total you received, you can't fool yourself.
Why plain ROI lies about time
The formula has no clock in it. That is its biggest weakness, and it's the easiest one to miss because the arithmetic looks complete.
Back to the opener with real numbers. Investment A returns 50% over five years. Investment B returns 20% in one year. On paper A wins by a mile. To compare them honestly you annualize each — convert it to a per-year rate:
Annualized ROI = ((1 + ROI)^(1 / years) − 1) × 100
For A: (1.50)^(1/5) − 1 = 8.4% per year. For B: 20% is already a one-year figure, so 20% per year.
B earns more than twice as fast. If you could repeat B, five years compounds to roughly 149% total — far more money than the 50% A ever delivered. The unannualized 50% was a mirage created by hiding the time axis.
This bites any time you compare things held for different lengths. A stock up 80% over eight years (about 7.6% a year) is a worse compounder than one up 15% in a single year — even though 80% is the number that gets quoted at parties.
ROI vs ROAS — different questions
Marketers lean on ROAS — return on ad spend — and it gets confused with ROI constantly. They answer different questions, and mixing them up loses real money.
ROAS = revenue / ad spend. Spend $1,000 on ads, pull in $4,000 in sales, and your ROAS is 4 (or 400%). It compares revenue to the ad bill and nothing else.
ROI compares profit to total cost. Take that same campaign: $1,000 in ads brought $4,000 in revenue, but the products themselves cost $3,200 to make and ship. Real profit is 4,000 − 3,200 − 1,000 = −$200. Total cost was $4,200.
ROI = (4,000 − 4,200) / 4,200 × 100 = −4.8%
A 4x ROAS that lost money. ROAS said "great," ROI said "you're underwater." ROAS is a fine top-line signal for a media buyer deciding which ad to keep running; it is not a measure of whether the business earned anything. Never hand someone a ROAS figure when they asked whether you made a profit.
What ROI still can't tell you
Even done correctly and properly annualized, ROI is a single number — and a single number flattens things that decide whether an investment was actually smart.
- Risk. A 12% annualized return from government bonds and a 12% return from a friend's crypto startup are not the same investment. One was nearly certain; the other could just as easily have been −100%. ROI reports the outcome, not the odds you were facing when you committed.
- Timing of cash flows. A project that pays your money back steadily each month is worth more than one that returns a single lump at the end, even at identical ROI, because early cash can be reinvested. ROI treats them as equal. Metrics like IRR exist precisely to fix this.
- Opportunity cost. A 6% return looks good if the alternative was 2% and bad if a boring index fund would have done 10% over the same window with less stress. ROI never measures itself against what you gave up.
- What you counted as "cost." Leave out your own labor, taxes, or transaction fees and the ROI inflates. A house flip advertising 25% ROI often ignores months of unpaid weekends and closing costs that would halve the real figure.
Two investments with the same ROI are almost never equally good. ROI ranks outcomes; it does not rank decisions.
Use it as a starting point, not a verdict
ROI is the right first calculation and the wrong last one. Run it, annualize it so time stops hiding, then ask the three questions the number quietly buries: How risky was this? When did the cash actually arrive? What else could the money have done?
The ROI Calculator handles the arithmetic — including the annualized version — right in your browser, with nothing uploaded and no sign-up, so you can run a dozen scenarios in a minute without handing your numbers to anyone. But it only produces a number. Whether that number means the investment was smart is a judgment it can't make for you. Line your options up apples-to-apples on the same time scale first, then let risk and opportunity cost break the tie. A smaller ROI you understand beats a bigger one you haven't questioned.