Sales are not money you keep
Here is the calculation almost everybody does. The campaign cost £1,000. It brought in £5,000 of sales. That is a 400% return, and it goes in the report.
Except £5,000 of sales cost you something to deliver. Materials, labour, subcontractors, the lot. On a 25% margin that £5,000 is worth £1,250 of actual money.
So you spent £1,000 and made £1,250. A profit of £250, and a return of 25%. Still positive, still worth doing, and nothing like 400%.
Where it goes properly wrong
The two numbers do not just differ in size. They can differ in sign.
Same £1,000 spend, £3,000 of sales, same 25% margin. On the sales figure that is a 200% return. In money it is £750 of margin against £1,000 of cost, which is a loss of £250.
That is the exact shape of a campaign that keeps getting repeated. The top line went up. Everybody was busy. There was less money at the end than at the start, and nothing in the reporting said so.
The cost of kit is not the sticker price
A £10,000 machine is rarely a £10,000 decision. What belongs in the number:
- Finance interest, if it is not being paid for outright.
- Delivery and installation.
- Training, including the trainer's time and the trainee's.
- The days the old way carried on while somebody learned the new one.
That last one is invisible and it is often the largest. It also lands entirely in the first year, which is the year the payback calculation cares about most, so leaving it out shortens the apparent payback by exactly the amount you left out.
ROI and payback are different questions
ROI asks: was this worth doing? Payback asks: how long is my money tied up?
For most small businesses the second matters more, because cash is what actually runs out. A 300% return over five years is a worse problem than a 40% return in four months if the money is needed in June. Both are on this page for that reason, so notice which one you instinctively quoted when you were justifying the purchase.
The resale value, and how much to trust it
Counting what the machine will be worth at the end is legitimate and it is the softest figure in the whole calculation. It is money you only see if you actually sell it, at a price you are guessing at, on a date you have not chosen.
Look at the answer with it and without it. If the decision only works with the resale value in, the decision depends on a second-hand market three years from now.
It measures against doing nothing, which is rarely the option
It measures against doing nothing. Doing nothing is rarely the real alternative.
The honest question is whether the same money spent on something else, or simply kept in the account against a quiet quarter, would have done better. No calculator answers that, and any that claims to is comparing your decision against an imaginary one.
Common questions
How do I work out ROI on a marketing campaign?
Not by dividing the sales it brought in by what it cost, which is what almost everybody does. Sales are not money you keep. Spend £1,000, generate £5,000 of sales on a 25% margin, and the actual return is £1,250, so you made £250 on £1,000 spent. That is a 25% return, not the 400% you get from dividing £5,000 by £1,000.
Why does that matter so much?
Because the two numbers can point in opposite directions. On the same £1,000 spend, £3,000 of sales at a 25% margin is £750 of margin against £1,000 of cost: a loss. Divide the sales by the cost and it looks like a 200% return. That is the exact shape of a campaign that keeps getting repeated, because the top line went up and everybody was busy and there was less money at the end.
What should I include in the cost of a machine?
More than the price on the invoice. Finance interest, delivery, installation, and the training nobody counts. Then the part almost everyone leaves out: the days the old way carried on while somebody learned the new one. All of that lands in the first year, which is the year the payback calculation actually cares about, so leaving it out makes the payback look shorter by exactly the amount you left out.
Is ROI or payback the more useful number?
For a small business, usually payback. ROI asks whether it was worth doing at all. Payback asks how long your money is tied up, and cash is the thing that runs out. A 300% return over five years is a worse problem than a 40% return in four months if you need the money in June. This page shows both because they answer different questions and people quote whichever flatters the decision they already made.
Should I count what the machine will be worth when I sell it?
You can, and treat it with suspicion. It is money you only see if you actually sell it, at a price you are guessing at, on a date you have not chosen. It belongs in the sum and it is the softest number in it, so it is worth looking at the answer with and without it before deciding anything.
It lost money. Was it a mistake?
Two different questions worth separating. Was it the wrong thing to buy, or the right thing bought too early? Kit that will pay back over three years and has done one year is not a mistake yet, it is a year in. A campaign that lost money on a fair margin figure is a clearer answer, because campaigns do not have three years to come good.
What is this comparison missing?
The next best thing you could have done with the money. This measures against doing nothing, and doing nothing is rarely the real alternative. The honest question is whether the same money spent elsewhere, or simply kept in the account for a quiet quarter, would have done better. That is a judgement rather than a calculation and no tool can make it for you.