How to Calculate Automation ROI Before You Build

Before You Sign Off on the Project

Somebody on your team (maybe you) wants to build an automation. A workflow to route leads, a bot to answer support questions, something that summarizes calls automatically. It sounds useful. It probably is useful. But nobody has actually written down what it's worth, and the quote from whoever's building it is sitting in your inbox waiting on a yes.

This happens constantly, and it's a big part of why automation projects fail before they launch. Plenty of automation projects get greenlit purely because they sound smart, well before anyone's run the actual numbers. To be fair, that's on us as an industry too, calculating automation ROI properly takes maybe 20 minutes and most people have never been shown how to do it. So teams either approve based on gut feel, or they kill a genuinely good project because it looks expensive on paper without ever comparing it to what the manual version already costs.

Think of this as a worksheet you run yourself before anyone quotes you a price. Grab a calculator, plug in your own numbers, and you'll know before you spend anything whether a given automation pays for itself in six weeks or takes eighteen months to break even.

The Three Numbers That Actually Matter

Every automation ROI calculation comes down to three inputs, and most people only think about the first one.

Time saved times loaded hourly cost. This is the obvious one. Somebody stops doing a manual task, so multiply the hours saved by what that person's time actually costs the business.

Error cost. Manual processes make mistakes. Wrong data entered, a follow-up that never happened, a duplicate record created, a wrong deal value logged. These have a real dollar cost even though almost nobody tracks it.

Opportunity cost. The revenue you're not capturing because a lead sat unanswered for six hours, or a client rescheduled and nobody rebooked them, or a call summary never made it into the CRM so the next rep started from zero.

Most ROI conversations stop at the first number. That's how you end up under-valuing automations that look expensive on a spreadsheet but are actually recovering revenue nobody was counting.

Step 1: Time Saved x Loaded Hourly Cost

Start with the task itself. Write down, honestly, how much time a person actually spends on it per day, rather than the polished estimate you'd give in a meeting. Time yourself for a few days if you're not sure, most people underestimate this by 30 to 50%.

Multiply daily minutes by the number of people doing the task, then by 5 (or however many working days apply), then by 52 weeks. Divide by 60 to get annual hours.

Next you need a loaded hourly cost, which goes beyond the wage alone. It includes payroll tax, benefits, and overhead. A rough shortcut a lot of small businesses use is wage times 1.25 to 1.4. If you don't want to calculate it precisely, $25 to $35/hour is a reasonable blended rate for admin, ops, or junior sales roles at most small service businesses, adjust up if your team earns more.

Annual hours saved times loaded hourly cost gives you your first number. For a task eating 45 minutes a day across two people at $30/hour loaded, that's roughly 1.5 hours a day, 390 hours a year, and about $11,700.

Step 2: The Cost of Errors You're Currently Absorbing

This one gets skipped constantly because it's harder to see. Manual processes fail quietly. Somebody forgets to log a callback. A deal value gets typed wrong. A lead gets entered twice and a rep works it twice, annoying the prospect.

You don't need a precise figure here, a defensible estimate works fine. Ask: how often does this specific error happen per week or per month, and what does it cost when it does? A missed follow-up on a lead worth $2,000 in average deal value, happening twice a month, is $48,000 a year in leads that simply evaporate. Even if you think that's high and cut it by 75%, you're still at $12,000. That's real money attached to a process most people describe as "eh, it happens sometimes."

If you genuinely don't know your error rate, that's fine, just track it for two weeks before you build anything and count instances as they come up. Most owners are surprised by the total once they actually write it down.

Step 3: Opportunity Cost, the Number Everyone Underrates

This is the hardest one to pin down and the one worth taking most seriously, because it's usually the biggest.

Across the businesses we've measured this for, the pattern holds up consistently: response within 5 minutes converts dramatically better than response after an hour, and conversion drops sharply the longer a lead waits. If your current process means leads sit for 2 to 4 hours before anyone replies (which is extremely common for small teams juggling other work), you're losing a real share of leads that a faster automated first response would have caught.

To estimate it, take your monthly lead volume, multiply by your current close rate, multiply by average deal value, and estimate what percentage of that volume is lost purely to slow response, separate from lead quality. Even a conservative 10 to 15% recoverable share on a modest lead volume adds up fast. A business getting 100 leads a month at a 20% close rate and a $1,500 average deal value is sitting on $30,000 a month in total pipeline value, recovering even 10% of that from faster response is $3,000 a month, or $36,000 a year.

Same logic applies to no-shows recaptured by automated reminders, or support tickets resolved instantly instead of sitting in a queue overnight while the customer looks elsewhere.

Step 4: Subtract What the Automation Actually Costs

Here's where a lot of ROI math gets dishonest, usually by omission rather than intent. The full cost of an automation isn't just the build fee.

Add up four things: the one-time build cost, the ongoing platform subscription (n8n, Make, HubSpot, Twilio usage, whatever's in the stack), a realistic maintenance allowance for when an API changes or a new edge case shows up, and a ramp period where a human is still double-checking output before you can trust it running unattended. Most automations aren't running at full reliability in week one, that tuning time costs something too even if nobody puts it on an invoice.

A typical n8n or Make workflow for something like lead routing or call summary logging runs a few hundred to low thousands of dollars to build depending on complexity, plus $20 to $100 a month in platform costs depending on volume. Add that total to whatever the maintenance and ramp allowance comes to, and you've got the number you'll divide the benefit total by in the next step.

Step 5: Payback Period and a Go/No-Go Threshold

Add your three benefit numbers together (time saved, error cost avoided, opportunity cost recovered), then divide the total build cost by that annual figure and multiply by 12. That's your payback period in months.

If payback lands under 3 months, it's an easy yes. Between 3 and 9 months, still generally worth it for anything that isn't a one-off task. Past 12 months, it's worth asking whether the task is actually recurring and stable enough to automate, or whether it changes too often for a build to hold its value. We tell clients the same thing we'd want told to us: if the math only works by using best-case numbers everywhere, don't build it yet. Use the conservative number and see if it still clears the bar.

A Worked Example: Lead Follow-Up Automation

Take a small service business getting 150 inbound leads a month across a website form and Facebook lead ads, currently followed up by hand within a few hours, sometimes longer.

Time saved: two people spending 40 minutes a day combined on manual follow-up, logging, and reminders. At $30/hour loaded, that's roughly $10,400 a year.

Error cost: leads occasionally fall through entirely, maybe 3 to 4 a month go completely unanswered. At a $1,200 average deal value and a 20% close rate, that's about $1,000 to $1,400 a month in dropped pipeline, call it $14,000 a year conservatively.

Opportunity cost: even among leads that do get followed up, response time averages 3 hours. Speeding that to under 5 minutes recovers maybe 8 to 10% additional close rate on the leads that were previously converting, roughly $8,000 to $10,000 a year on top of the error cost above.

Total annual benefit: somewhere around $32,000 to $34,000, using conservative assumptions throughout, in the same range as what a small business can typically expect to save by automating a process like this.

What we'd actually build: a webhook in n8n (or Make, either works fine here) that catches every new lead the second it hits the form or Facebook ad, writes it into HubSpot with source and priority tagged automatically, and triggers an instant WhatsApp or SMS via Twilio confirming receipt and setting expectations. A day 1, day 3, day 7 follow-up sequence runs automatically for anyone who hasn't responded, and anything urgent gets flagged to a rep immediately instead of sitting in a queue. A Google Sheets log tracks every lead's path so you can see volume and outcome without digging through HubSpot.

Build cost for something like this typically runs $1,500 to $3,000 depending on how many branches and integrations are needed, plus maybe $50 to $80 a month across the platforms. Against $32,000 in annual benefit, payback lands under two months.

Yes, Some of This Is Guesswork

Fair pushback, and we'd raise it too if we were on the buying side. The response is that every number in this worksheet is meant to be your own, pulled from your own time tracking and your own close rates. Whoever hands you a generic industry figure to make a sale look good is doing the opposite of this exercise, which is exactly the point of running it yourself before anyone quotes you a price.

The other honest answer is that even a rough estimate beats no estimate. Nobody's calculating time-saved-times-hourly-cost to the penny, and that's fine, the goal is directional accuracy. If your conservative low-end number still shows payback inside 6 to 9 months, you don't need precision to justify moving forward. If the numbers only clear the bar once you stack generous assumptions on top of each other, that's useful information too, it tells you the project sits closer to a maybe than a confident yes.

And if the math genuinely doesn't clear, that's a legitimate outcome of this exercise. Some tasks aren't worth automating yet, especially low volume, highly variable processes, or ones that change shape every few weeks, and it's better to find that out with a worksheet than after paying for a build.

Run the Numbers Before You Commit Budget

If you want a second set of eyes on this before you commit budget, that's what the free audit is for. We'll sit down, go through your actual volume, your actual time spent, and your actual close rates, and tell you honestly whether the automation you're considering clears the payback bar, and roughly how fast.

Book a free 60-minute call and we'll walk through the math on your specific process before you spend a dollar building anything.