How to Calculate Automation ROI for a Small Business (Simple Formula and Worksheet)

Updated 2026-10-30 · AI Agent Starter Kit

Every automation tool promises to "save hours a week." Some do. Many save a few minutes and cost an afternoon of setup and a monthly fee. Before you build or buy an automation, it's worth spending 20 minutes working out whether it will pay for itself, and how fast.

This guide gives you a simple formula, a way to measure the task honestly, what to count on the cost side, how to value faster replies (not just time saved), a copy-paste Google Sheets worksheet, an illustrative example and the mistakes that make automation ROI look better on paper than in real life.

The year-end planning season is a good time for this. If you're building next year's budget, the worksheet tells you which tools deserve a line in it.

The short formula

For a small business, automation ROI comes down to three numbers:

Monthly benefit = (hours saved per month x value of your hour) + extra revenue from doing the task better

Monthly cost = subscription and running costs + upkeep time x value of your hour

Payback period (months) = one-time setup cost / (monthly benefit - monthly cost)

If the monthly benefit is lower than the monthly cost, the automation never pays back, however clever it is. If it pays back in a few months, it's usually worth doing.

ROI as a percentage, if you want it for a year:

Annual ROI % = (12 x (monthly benefit - monthly cost) - setup cost) / (setup cost + 12 x monthly cost) x 100

Step 1: Pick one task and describe it precisely

"Email" is not a task. "Writing first replies to new website inquiries" is. Pick one task and write down:

Question	Example answer
What triggers it?	A new contact form submission
What do you do, step by step?	Read it, check the calendar, write a reply, log it in the sheet
How often does it happen?	Count it (see Step 2)
How long does each one take?	Time it (see Step 2)
What goes wrong when it's slow or missed?	Lead books someone else
Which part needs your judgment?	Price and dates

The last row matters. The parts that need your judgment usually stay with you, so they don't count as time saved.

Step 2: Measure the task honestly for two weeks

Guesses about time are usually wrong. For two weeks, keep a tally:

TASK LOG: [task name]

Date | Count | Minutes total | Notes (interruptions, rework)
     |       |               |
     |       |               |

Two-week total: count ____  minutes ____
Average minutes per item: ____
Items per month (x 2.17): ____

Include switching time. If answering one inquiry takes 4 minutes but pulls you out of a job for 15, that matters. Note it in the notes column and decide whether to count it.

Step 3: Work out what part of the task is actually removed

Most small business automations don't remove a task entirely. They remove part of it. Be specific:

Type of automation	What it removes	What you still do
Logging leads to a sheet	Copying details by hand	Checking the row is right
Drafting replies	Writing from a blank page	Reading, editing, sending
Sorting and labeling email	Deciding where each email goes	Spot-checking labels
Weekly report	Counting numbers by hand	Reading it, deciding what to do
Reminders and follow-up drafts	Remembering who to chase	Approving each message

A fair estimate: time saved per item = old minutes per item - new minutes per item (review time included). Time the new version for a week after setup rather than assuming it drops to zero.

Step 4: Put a value on your hour

Use one of these, and write down which one:

What you'd pay someone else to do the task (useful for admin work).
What you bill clients per hour (only if you'd actually fill the saved time with paid work).
What your time is worth to you (useful for evenings and weekends you'd get back).

Be honest. If the saved time won't turn into paid work, using your billing rate overstates the benefit.

Step 5: Count the revenue side (carefully)

Some automations matter less for the time they save and more for doing the task faster or more reliably. Faster first replies to leads and follow-ups that don't get forgotten are the usual examples.

Use your own numbers, not industry claims:

Extra revenue per month = extra jobs won per month x average job value x your margin

To estimate "extra jobs won", look at your own data: how many leads went cold last quarter because nobody replied or followed up? If you don't track it, start now and leave this line at zero until you have a few months of data. The close rate calculator and the lead response calculator help you see what changes in close rate or reply speed would mean, and both label their results as estimates.

Step 6: Count every cost

Cost	One-time or monthly	Notes
Setup time	One-time	Your hours x your hourly value, including testing
Templates or kits	One-time	Purchase price
Automation platform	Monthly	Check current pricing on the vendor's site
AI usage (API)	Monthly	Usually small for low volume, but check your provider's pricing
Other software it needs	Monthly	Only if you don't already pay for it
Upkeep	Monthly	Fixing breaks, updating prompts, checking output
Review time	Per item	Already counted in Step 3

Upkeep is the line most people skip. Budget at least some time each month, especially in the first few months.

The automation ROI worksheet (copy into Google Sheets)

Put labels in column A and values in column B:

Row 1:  Task                              | [name]
Row 2:  Items per month                   | 
Row 3:  Minutes per item before           | 
Row 4:  Minutes per item after (review)   | 
Row 5:  Hours saved per month             | =B2*(B3-B4)/60
Row 6:  Value of one hour ($)             | 
Row 7:  Time value per month ($)          | =B5*B6
Row 8:  Extra jobs won per month          | (0 until you have data)
Row 9:  Average job value ($)             | 
Row 10: Margin (as %)                     | 
Row 11: Extra profit per month ($)        | =B8*B9*B10
Row 12: MONTHLY BENEFIT ($)               | =B7+B11
Row 13: Subscriptions per month ($)       | 
Row 14: AI usage per month ($)            | 
Row 15: Upkeep hours per month            | 
Row 16: MONTHLY COST ($)                  | =B13+B14+B15*B6
Row 17: Setup hours                       | 
Row 18: Purchase price ($)                | 
Row 19: ONE-TIME COST ($)                 | =B17*B6+B18
Row 20: NET PER MONTH ($)                 | =B12-B16
Row 21: PAYBACK (months)                  | =IF(B20<=0,"Never",B19/B20)
Row 22: FIRST-YEAR ROI %                  | =IF(B19+12*B16=0,"",(12*B20-B19)/(B19+12*B16))

Format B22 as a percentage. Copy the sheet once per automation you're considering, then compare the payback rows side by side.

How to read the result

Payback	What it usually means
"Never"	Monthly cost is higher than the benefit. Skip it or simplify it
Under 3 months	Strong case, if your measurements are honest
3 to 12 months	Reasonable, especially if the task will keep growing
Over 12 months	Only worth it for reliability reasons (things currently get missed)

These bands are rules of thumb for a small business, not standards. A long payback can still be worth it if the task is one you dread or one where mistakes cost you clients.

An illustrative example

This example uses made-up numbers to show the method, not a real result. A one-person bookkeeping practice is deciding whether to automate first replies to new inquiries.

The two-week log shows about 6 inquiries a week (about 26 a month), each taking about 12 minutes to read, check availability and reply. After a test week with an automation that logs each inquiry and drafts the first reply, the owner spends about 4 minutes per inquiry reviewing and sending. That's 26 x 8 minutes, or about 3.5 hours a month. The owner values admin time at what a part-time assistant would cost and gets a modest monthly time value.

Running costs are small: the owner self-hosts the automation platform on an existing server, and AI usage at this volume costs a few dollars a month. Upkeep is budgeted at 1 hour a month. Setup took about 3 hours plus the purchase price of a template.

The worksheet shows payback in under three months from time alone. The owner leaves "extra jobs won" at zero for now, but starts tracking how many inquiries got a reply within an hour, so the revenue line can be filled in with real numbers after a quarter.

Common mistakes

Counting the whole task as saved. You still review, approve and fix. Time the new version.
Using your billing rate for admin time you wouldn't bill anyway.
Ignoring upkeep. Automations break when a form field or label changes.
Inventing the revenue line. Leave it at zero until you have your own data.
Automating a rare task. Something you do twice a month rarely pays back.
Automating a messy process. Fix the steps first, or you automate the mess. The guide on how to systemize your small business helps.
Letting automation send for you to save the review time. The review is what keeps mistakes away from clients.

Using AI for the calculation (carefully)

AI can help you list the steps of a task or turn your worksheet into a short summary. It should not supply the numbers:

Every number comes from your log or your bills. If an AI says "automation saves small businesses X hours a week", ignore it for your worksheet.
Formulas do the math. Let the sheet calculate payback and ROI.
Check pricing at the source. AI often gives out-of-date subscription prices.
Automations that use AI should draft, not send. Keep a human review step in anything that writes to clients or leads, and treat incoming email text as data, not instructions.

For ideas on what to measure first, see small business automation ideas and how to automate admin tasks.

Test the numbers with real workflows

If you want to fill in the "after" column with real measurements instead of guesses, the Solo Ops Command Center bundle gives you three n8n workflow kits to try on your own Gmail and Google Sheets: Inbox Chief of Staff (sorts Gmail into labels, drafts replies in your voice, sends a 6 PM digest of what still needs you), Speed-to-Lead (logs every new inquiry to a sheet and drafts a first reply and Day 2, 5 and 10 follow-ups that stop when the person replies) and the Weekly CEO Report (a Monday summary of new leads, replies, stuck leads and overdue emails, counted by code and saved as a draft to you), plus a dashboard sheet that charts the weekly trend. The weekly counts give you real "items per month" numbers for the worksheet. To be clear about what it doesn't do: it doesn't calculate ROI for you, time your tasks, host n8n or pay for AI usage (you bring your own n8n and AI key), and it doesn't send anything. Every reply waits in Gmail Drafts. Nothing sends without you. Get the bundle

Want to try a workflow before spending anything? The free Morning Briefing Agent is a single n8n workflow that emails you your meetings, the 3 unread threads most likely to matter and one priority every weekday at 7 AM. It's read-only on your calendar and inbox and the only email it sends goes to you. It's a low-risk way to see how long setup and upkeep take for you before you fill in the cost rows. Get it free

Related: how to follow up with leads during the holidays (a task where reliability matters more than minutes saved).

Bundle: best value

Solo Ops Command Center

All three paid kits (Inbox Chief of Staff, Speed-to-Lead, Weekly CEO Report) plus a bundle-only dashboard sheet. $87 separately.

Get the bundle ($69)

Free template

Morning Briefing Agent

One calm 7 AM email: your meetings, the 3 emails that matter, and one priority. Read-only. 10-minute setup.

Get it free

Free weekly email

One tested AI workflow a week

Short, practical n8n and Gmail automations for one-person businesses. Drafts-only, human-in-the-loop, no hype. Unsubscribe anytime.

Subscribe free