ROI calculator: How to work out return on investment

Work out the return on any investment: a campaign, a hire, a piece of software, a property, or a whole company. Enter what you put in and what you got back, and this calculator gives you the return as a percentage and as a cash figure.

It also runs in reverse. If you know your budget and the return you need, it tells you what the investment has to pay back. And if you add a time period, it converts the total return into an annualized figure, which is the only fair way to compare a deal that took nine months against one that took four years.

ROI is the ratio of what you made to what you spent. Nothing more.

Both halves are cash. The gain is what the investment returned above its cost, and the cost is everything you put in to get it. Express the ratio as a percentage and you have a number that compares a $4,000 ad test against a $40m acquisition on the same scale. That comparability is why ROI outlasted every metric invented to replace it.

The trap is the name. "Return on investment" gets used loosely for at least four different calculations: return on invested capital, return on equity, average rate of return, and earnings per share. They answer different questions and produce different numbers from the same company. When someone hands you an ROI figure, the useful question is not whether it is high but what went into the denominator.

This calculator uses the simple version, which is the one nearly everyone means: net gain divided by total cost.

The formula needs two inputs:

ROI = (Amount returned − Amount invested) ÷ Amount invested

Or, if you already know the gain in cash:

ROI = Gain ÷ Amount invested

Multiply by 100 for a percentage. A result of 0.4 is an ROI of 40%.

What the formula leaves out matters as much as what it includes. There is no term for risk, no term for time, and no term for the money you could have made doing something else. ROI tells you the size of the return and stays silent on whether it was worth taking.

A paid acquisition campaign. You spend $250,000 on a new demand generation program. Over the following two years it drives $200,000 of incremental gross profit each year, so the total return is $400,000.

ROI = ($400,000 − $250,000) ÷ $250,000 = 0.60 = 60%

Note the choice made in that denominator. If the $250,000 is media spend only, and the two people running the program cost another $300,000 in salary, the honest ROI is negative. Fully loaded cost is where marketing ROI claims usually fall apart.

A software purchase. A 40-person support team buys a tool for $60,000 a year. It removes about 6 hours per person per week. At a $35 fully loaded hourly cost, that is 40 × 6 × 52 × $35, or $436,800 of recovered capacity.

ROI = ($436,800 − $60,000) ÷ $60,000 = 6.28 = 628%

Treat that number with suspicion. Recovered hours are only a return if you actually convert them into output or headcount you did not hire. Time saved that dissolves into the working day is not $436,800.

An acquisition. You buy a competitor for $600,000 and sell the combined business three years later, attributing $900,000 of the proceeds to that unit.

ROI = ($900,000 − $600,000) ÷ $600,000 = 0.50 = 50%

Fifty percent sounds strong. Spread across three years it is 14.5% a year, which is a different conversation, and the next section is why.

Two investments both return 50%. One takes a year, the other takes four. They are not the same investment, and ROI on its own cannot tell them apart.

Annualized ROI fixes this by asking what constant yearly rate would compound to the same total return:

Annualized ROI = (1 + ROI)^(1 ÷ years) − 1

That 50% over three years works out to 14.5% a year. Over one year it stays 50%. Over ten it falls to 4.1%, which is worse than leaving the money in treasuries.

Open the time period section in the calculator above and it will do this for you, from either a length of time or a pair of dates. Use it whenever you are comparing two options, because comparing raw ROI across different holding periods is the single most common way the metric gets misread.

Positive is not the same as good. A positive ROI only means you got back more than you put in, and it says nothing about whether the capital would have done better somewhere else.

The professional version of the question uses a hurdle: a minimum return below which the investment is not worth doing. Three common ones, in ascending order of rigour:

For fast-cycle marketing and sales spend, most teams pair ROI with a payback period instead, since a campaign that returns 30% in six weeks and one that returns 30% in six quarters are not remotely comparable investments.

Four metrics, constantly confused, each answering a different question.

Two traps worth naming. ROAS uses revenue, not profit, so a 4x ROAS on a product with a 20% gross margin is a loss. And ROI includes all costs while ROAS includes only media, which is why the same campaign can show a triumphant ROAS and a negative ROI. If you are reporting to a finance audience, they mean ROI, and they mean fully loaded.

Return on equity is the other frequent mix-up. ROE divides by shareholders' equity alone, while ROI divides by everything invested including debt. A leveraged business can post a spectacular ROE and a mediocre ROI at the same time.

It ignores time. Covered above, and it is the big one.

It ignores risk. A guaranteed 12% and a 12% that came from a coin flip are identical to the formula. They are not identical decisions.

It is easy to manipulate. Both the numerator and the denominator are matters of judgment. Move some costs out of the denominator, extend the window over which you count the gain, and a weak program starts to look strong. This is not usually fraud, it is just optimism, but it is why an ROI figure should always come with its inputs attached.

It struggles with attribution. The formula assumes the gain was caused by the investment. Revenue that would have arrived anyway, or that came from a different campaign, inflates the result. Anything with a long or shared path to purchase is exposed to this.

It says nothing about scale. A 300% ROI on $5,000 returns $15,000. A 30% ROI on $5m returns $1.5m. Percentages hide the size of the prize, so read ROI and absolute gain together. The calculator above shows both for exactly this reason.

Only two levers exist: raise the return or cut the cost. In practice the cost side moves faster.

Subtract the amount invested from the amount returned, divide by the amount invested, and multiply by 100. Put $10,000 in, get $14,000 back, and the ROI is ($14,000 − $10,000) ÷ $10,000 × 100 = 40%.

It depends on what the money costs you and what else it could have done. As a floor, anything material should clear your weighted average cost of capital, which for most private companies sits somewhere between 8% and 15%. Marketing and sales spend is usually held to a category benchmark and a payback period instead of a single ROI threshold.

You got back 30% more than you put in. Invest $100 and you finish with $130, of which $30 is gain. It does not say how long that took, which is why annualized ROI matters.

ROI divides net gain by total cost. ROAS divides revenue by ad spend alone. ROAS ignores both the cost of goods and every cost outside media, so it always reads higher. A 4x ROAS on a 20% gross margin product is losing money.

Yes. Any time you get back less than you put in, the ROI is negative, and it bottoms out at −100% when you lose the entire investment. The calculator handles negative returns.

Subtract the purchase price and all costs of the purchase from the sale proceeds, then divide by that total cost. Include stamp duty, fees, and improvements, or the figure will flatter the deal. For a property held several years, use the annualized ROI.

The period the money was actually committed, from the day it went out to the day it came back. If you use a date range, the calculator converts it for you.