All free tools

Free tool · No signup

Automation ROI Calculator

See how many hours and dollars automating your team's repetitive work saves, and how fast it pays back.

Tasks your team repeats

Times per week is per person. Hourly cost is the loaded rate: wage plus taxes, benefits and overhead, often 1.25–1.4× the wage.

Why 70% automatable, not 100%?

Real tasks have exceptions: a blurry scan, a lead with no website, a customer who needs a person. Someone still handles those and checks the output. 70% is a conservative default. Go higher only for fully digital, rule-based work.

  1. Mistakes (optional)
  2. Mistakes (optional)
Or add an example:
Project cost

What it costs

Typical starting points. Replace them with a real quote when you have one.

Hosting or plan fees, AI API usage and time for fixes.

Payback

3.2 months

Saved per month

$476

$576 − $100 running

Hours saved per month

14 h

170 h a year · 0.08 FTE

Net after 12 months

$4,211

Net after 36 months

$15,632

$0$2K$4K$6K$8K024681012Break-even
Months from launch. Cumulative savings overtake cumulative cost at month 3.2. After 12 months, savings total $6,911 against $2,700 in cost.

Automate these first

  1. 1. Copy invoices into accounting$303/mo
  2. 2. Build the weekly report$273/mo

Payback in 3.2 months. I can scope the top task for a fixed price.

Top task: Copy invoices into accounting, $303 a month.

Your task list goes into the message so you don't have to type it again. You can edit it before sending.

How this is calculated
  • Hours saved per month = times per week × 4.33 × minutes ÷ 60 × people × automatable %
  • Monthly savings = hours saved × hourly cost + runs × mistake rate × automatable % × cost to fix
  • Net per month = $576 savings − $100 running = $476
  • Payback = $1,500 build ÷ $476 net per month = 3.2 months
  • Net after N months = net per month × N − build cost. FTE = hours ÷ 173.2 (40 h × 4.33 weeks).

No hidden multipliers. The estimate assumes the task keeps the same volume and that saved time goes to other useful work. Amounts work in any currency, as long as they're all in the same one.

How to calculate automation ROI honestly

The basic sum is simple: hours saved × what those hours cost, minus what the automation costs to build and run. Most ROI figures go wrong in the inputs, not the arithmetic. Three things keep them honest:

  • Use the loaded hourly cost. An hour of someone's time costs more than their wage once you add payroll taxes, benefits, equipment and office space. 1.25 to 1.4 times the wage is a common rule of thumb. Using the bare wage understates the savings; using a consultant's rate overstates them.
  • Assume partial automation. Real tasks have exceptions: an invoice in an odd format, a lead with no website, an email that needs a person. Someone handles those and checks the output. That's why the calculator defaults to 70% automatable rather than 100%.
  • Count the running cost. Hosting or a plan subscription, AI model usage, and a few hours a quarter for fixes when a connected tool changes. An automation that saves $150 a month but costs $200 to run never pays back, and the calculator says so.

The formula is shown under the result, with your numbers in it. There are no hidden multipliers: if a number looks off, you can see which input caused it.

Which tasks are worth automating

The best candidates share three traits:

  • Rule-based, or close to it. The steps are the same each time, or the judgment involved is the kind an AI model handles well: classifying, extracting fields, drafting a first reply.
  • High volume. A 3-minute task done 200 times a week is worth more than a 2-hour task done once a month. Volume is what makes the build cost pay back.
  • Digital inputs. Emails, forms, PDFs, spreadsheets and anything with an API. Paper, phone calls and screens without an export take more work first.

Typical examples are invoice processing, enriching new leads, building a weekly report, triaging a shared inbox and typing data from PDFs into another system. Each is one click to add as an example above. A process that still changes every week is a poor candidate: settle it first, then automate it.

No-code or custom code?

No-code tools like Zapier, Make and n8n cost less to build and are easy to change, but plan fees grow with volume and complex logic gets hard to maintain. Custom code costs more up front, runs cheaply at high volume and handles messy edge cases better. Try both presets in the calculator: for low-volume tasks no-code usually pays back faster, and at high volume custom code often wins over 36 months.

For a side-by-side example, see n8n vs custom Python for AI workflow automation, which builds the same lead-intake workflow both ways and compares maintenance.

Chatbot ROI

Switch to Support chatbot to estimate a chatbot instead. It counts two kinds of value: support conversations the bot answers without a person (at your cost per ticket), and extra leads it captures, for example from visitors who arrive after hours. Be conservative with the share it can answer: 20 to 40% of conversations is realistic for a bot that answers from your own docs, and only if those docs are current.

What the result doesn't include

  • Changes in volume. The estimate assumes the task stays the same size. If volume grows, savings grow with it.
  • Your team's time during the build. Budget a few hours for explaining the process and testing the first version.
  • Softer gains such as faster replies, fewer missed leads beyond the ones you enter, and less tedious work. They're real, but they're left out so the number stays defensible.

When you have a task in mind, see what AI workflow automation or a custom AI chatbot involves, or send the result and get a fixed quote.

By Muhammad Saif · Updated

Questions

What's a good payback period for automation?

Under 12 months is a comfortable yes for most small businesses, and simple automations of high-volume tasks often pay back in two to six. Past 18 to 24 months, the process is likely to change before the build has paid for itself, so start with a smaller or no-code version.

What does automation cost to maintain?

Budget for hosting or a plan subscription, AI model usage where it's used, and a few hours a quarter for fixes when a connected tool changes its API or a form changes. Put all of that into the monthly running cost; for a typical small workflow it's somewhere around $50 to $300 a month.

Will automation replace staff?

Usually not. It takes the copying, checking and re-typing off people's plates, and they spend that time on the work that needs judgment. The hours saved are capacity: what it's worth depends on whether that time goes to sales, service or growth without hiring.

Why not assume 100% of a task is automated?

Because almost every real task has exceptions: a scanned invoice the model can't read, a lead with no website, a customer who needs a person. Someone still handles those and checks the output. 70% is a conservative default; go higher only for fully digital, rule-based tasks.

What hourly cost should I use?

The loaded cost, not just the wage: add payroll taxes, benefits, equipment and office costs. A common rule of thumb is 1.25 to 1.4 times the hourly wage. For your own time as an owner, use what an hour of your time earns the business.

Is anything I enter stored?

No. The calculator runs entirely in your browser. Your numbers are kept in this browser and in the share link, and never sent to a server.