How to Calculate the ROI of Process Automation (Before You Buy Any Tool)

Automation software is easier to buy than ever. Zapier, Make, n8n, Power Automate and dozens of other tools promise to save hours every week. Some automations deliver that and more. Others take weeks to build, break every month and save almost nothing. The difference is usually decided before anyone opens an automation tool: it comes down to picking the right processes.

This guide gives you a simple way to estimate the return on automating a task, so you can prioritize with numbers instead of gut feel.

Why this matters: how much time goes to repetitive work

The research on this is remarkably consistent. It suggests a large share of everyday work is repetitive, coordinative or both:

• In its 2017 report A Future That Works, the McKinsey Global Institute estimated that about half of the activities people are paid to do globally could technically be automated using technologies that had already been demonstrated, and that about 60% of occupations had at least 30% of their activities that could be automated. Very few whole jobs (under 5%) were fully automatable. In other words, automation mostly changes tasks, not whole roles. (McKinsey Global Institute, 2017)
• In a 2017 survey by Smartsheet, 42% of workers reported spending more than 10 hours a week, more than a quarter of their workweek, on repetitive, manual tasks, and 86% said automation makes workers more efficient and productive. (Smartsheet, 2017)
• Asana's Anatomy of Work Index 2021, based on a survey of more than 13,000 knowledge workers, found people spent 60% of their day on "work about work," such as chasing status updates, searching for information and switching between apps, rather than the skilled work they were hired to do. (Asana, 2021)
• Salesforce's 2026 State of Sales report, a survey of 4,050 sales professionals conducted in August and September 2025, found the average seller spends just 40% of their time actually selling. (Salesforce, 2026)

These studies measure different things in different ways, so don't treat them as a precise forecast for your business. But the direction is clear: in most organizations, there's meaningful time to win back.

The basic ROI formula

You don't need a spreadsheet model with 40 tabs. For most automation decisions, this is enough:

Annual benefit = time saved + errors avoided + speed gained

Annual cost = build effort + tool subscriptions + ongoing maintenance

ROI = (annual benefit − annual cost) ÷ annual cost

And the number most owners care about:

Payback period = upfront cost ÷ monthly benefit

Let's break down each part.

1. Time saved

This is usually the biggest and easiest benefit to measure.

Hours saved per year = (minutes per task × times per year) ÷ 60 × share of the task that is actually automated

The last factor matters. Few processes are 100% automatable. If someone still needs to review exceptions, count only the portion the automation takes over.

To turn hours into money, multiply by the fully loaded hourly cost of the people doing the work: their pay plus benefits, taxes and overhead, not just their wage.

2. Errors avoided

Manual data entry creates mistakes: a mistyped email address, an invoice sent to the wrong contact, a lead that never gets followed up. Estimate:

Error cost per year = errors per month × 12 × average time (or money) to fix each one

Don't forget harder-to-count costs, such as a lost deal because a lead sat untouched for a week. If you can't quantify these honestly, list them separately as "upside" rather than inflating the numbers.

3. Speed gained

Some automations matter because they make things faster, not just cheaper: instant lead routing, same-day onboarding, real-time inventory updates. If faster response plausibly wins more business, note it. Again, be conservative.

4. Costs

• Build: your time or a consultant's time to design, build and test.
• Tools: subscriptions for the automation platform and any additional app tiers needed for API access.
• Maintenance: apps change, APIs get updated and processes evolve. Budget some hours each year for upkeep. A common mistake is assuming automations run forever untouched.

A worked example (illustrative)

Here's a hypothetical scenario to show the math. The numbers are invented for illustration, so plug in your own.

A 10-person agency copies every new web inquiry from email into its CRM, creates a follow-up task and sends a "thanks, we'll be in touch" email.

• The process takes about 6 minutes per inquiry.
• They get about 15 inquiries a week, or around 780 a year.
• An automation could handle roughly 90% of the work, with someone still reviewing odd cases.

Hours saved per year = (6 × 780) ÷ 60 × 0.9 ≈ 70 hours

Now suppose about 4 inquiries a month are currently mistyped or missed, and each costs 30 minutes to sort out. That's another 24 hours a year, before counting any lost deals.

So the automation saves around 94 hours a year. Multiply by your loaded hourly cost to get the annual benefit, subtract tool and maintenance costs, then compare to the build effort. Simple automations like this often pay back quickly because they run hundreds of times a year.

Contrast that with a complex process that runs once a month and takes an hour. Even fully automated, it saves just 12 hours a year, and if it has lots of exceptions, it may cost more to build and maintain than it ever saves.

A quick scoring method for prioritizing

When you have a long list of candidates, score each from 1 to 5 on:

1. Frequency: how often does it happen?
2. Time per run: how long does it take manually?
3. Error impact: how costly are mistakes?
4. Rule clarity: are the steps consistent and rule-based, or full of judgment calls?
5. System access: do the apps involved have APIs or reliable integrations?

Then subtract a score for risk (what happens if it fails?) and complexity. Start with the tasks that score high on frequency, time and clarity, and low on risk. These are your quick wins, and they build confidence for bigger projects.

Signs a process is not ready to automate

• People do it differently every time. Standardize first, then automate.
• The underlying data is messy. Automation will just move bad data faster. (See our guide to CRM data hygiene.)
• It changes constantly. Wait until it stabilizes.
• It depends on judgment that's hard to describe. Consider AI-assisted steps with human review, or leave it manual.

Measure after launch

Before you build, write down the baseline: how long the process takes, how often it runs and how many errors occur. After launch, check those numbers again at 30 and 90 days. This keeps your ROI estimates honest and tells you which type of automation to invest in next.

The bottom line

Good automation ROI comes from high-frequency, rule-based tasks that touch systems with good integrations. Start small, measure honestly, include maintenance in your costs and expand from proven wins. If your list includes judgment-heavy steps, read our guide to practical AI adoption, since some of those tasks are now within reach too.