Automation ROI Calculator

How to Calculate ROI

ROI is the return an investment produces relative to what it cost:

ROI (%) = (Net Profit ÷ Investment Cost) × 100

Where Net Profit = Total Returns − Total Investment Cost.

The formula is trivial. Getting an honest answer out of it is not, because both inputs are easier to guess wrong than to measure. Almost every disappointing automation project had a defensible-looking ROI figure attached to it at the start.

A Worked Example

Take a three-person sales team. Each person spends roughly six hours a week on manual lead entry, chasing and follow-up admin — eighteen hours a week across the team, or 936 hours a year.

At a fully loaded cost of £22 an hour, that is £20,592 a year of labour going into work that a webhook, a CRM and a follow-up sequence can largely do on their own.

Now the investment side. A build like this typically comes to a one-off cost of around £4,500, plus a platform subscription of around £85 a month (check the live figures on Make.com's pricing page or in the n8n documentation if you are self-hosting, since both move). First-year investment: £5,520.

Here is the part most calculations skip. Automation does not remove all of the time. Exceptions still need a human, someone still checks the queue, and the first month is spent tightening data quality. Assume 70% of the hours are genuinely eliminated rather than 100%:

• Annual return: 936 × £22 × 0.70 = £14,414
• Net profit, year one: £14,414 − £5,520 = £8,894
• ROI, year one: (£8,894 ÷ £5,520) × 100 = 161%
• Payback period: £5,520 ÷ (£14,414 ÷ 12) = 4.6 months

Year two is where it gets interesting. The £4,500 build cost does not recur, so the investment drops to £1,020 of subscription and the same £14,414 return produces an ROI over 1,300%. This is why automation ROI should always be stated with a time horizon attached. A first-year figure understates a good project and a five-year figure flatters a bad one.

What a Good ROI Looks Like

• Under 0% — losing money. Revisit the thesis rather than the execution.
• 0–20% — marginal. The number is inside the error bars of your own estimate.
• 20–100% — solid, and typical of most sensible business investments.
• Over 100% — excellent, and worth a second look at your assumptions before you present it.

ROI vs Payback Period

ROI tells you how much you make relative to what you spent. Payback tells you how fast you get your money back. They answer different questions and they disagree more often than people expect.

A 200% ROI spread over five years is worse for most small businesses than a 50% ROI that pays back in six months, because the second one frees the capital to be used again. If cash flow is tight, weight the payback period and treat ROI as the tiebreaker.

Where Automation ROI Calculations Usually Go Wrong

Assuming 100% time elimination. The single most common error. Something always needs a human. Model 60–80% and be pleasantly surprised.

Ignoring the build and the second pass. Nearly every automation needs revisiting a few weeks after launch, once real edge cases have surfaced. That is real cost and it belongs in the investment side.

Counting saved hours as saved money. Hours only become money if the person does something more valuable with them or the headcount changes. If neither happens, the return is capacity rather than cash. Both are legitimate, but say which one you mean.

Forgetting the metered costs. Platform subscriptions are the visible number. Per-message, per-minute and per-token charges sit underneath and scale with volume, which is exactly when the ROI case matters most.

Frequently Asked Questions

Should I include my own time in the investment cost?

Yes, if you would otherwise have spent that time on billable or revenue-generating work. Time spent specifying, testing and fixing an automation is real cost even when no invoice is attached to it. Businesses that leave it out routinely underestimate a build by a third.

How do I put a number on fewer errors or faster response times?

Where you can, tie it to something observable: the cost of reworking one mis-keyed order, or the conversion rate difference between leads contacted within five minutes and leads contacted the next day. Where you cannot, list the benefit separately in words rather than inventing a figure for it. A clean ROI with a short qualitative list beside it is more persuasive than a padded one.

What time horizon should I use?

One year for the decision, three years for the business case. One year keeps you honest about build cost. Three years shows what the investment actually looks like once the one-off spend has washed through, which is the number that matters if the automation is going to keep running.

What if the ROI comes out negative?

Then the project may still be worth doing for reasons ROI does not capture, such as removing key-person risk or making a process survivable when someone is off. Just do not dress those up as a financial return. Decide on the real grounds.

Related Calculators

The Project Cost Calculator estimates the investment side for automation builds specifically, and the Employee Time Waste Calculator helps quantify the return side by putting hours against the manual work you are currently doing. If the project involves an AI voice agent, the AI Agent Cost Calculator covers the metered per-minute costs this general calculator cannot see.

For the wider context on what is worth automating first and why payback varies so much between processes, What Is Workflow Automation covers the ground properly.

Want a Custom ROI Analysis?

Book a free call and I will model the ROI for your specific automation or AI project, including the soft benefits and the metered costs a general calculator cannot know about. Bring your current process and roughly how many hours a week go into it.