Find the work worth automating, and what it is worth.

Pick the tasks your team does every week, tell us how often and how long they take, and answer two questions about each. You get the hours and money at stake, which tasks are quick wins, which need a bigger project, which to leave alone, and a 90-day plan.

Built from your own numbers
Tells you what not to automate
No sign-up, about five minutes

1
Your businessIndustry, hourly cost, how open your systems are
2
Your weekly tasksTick what applies, adjust volume and time
3
Your planRanked opportunities, payback and a 90-day roadmap

Your automation plan

Where to start

Value against ease

Bubble size shows hours saved.

Quick win
Bigger project
Small, easy
Not worth it yet

Ranked opportunities

Highest value you can realistically capture first.

Your first 90 days

Before you start

Want someone to check this against your real systems?

Send us the plan. An engineer will look at your top three tasks, the systems involved, and tell you honestly which would pay back and how we would approach them.

Review my plan

What makes a task worth automating

Four things have to be true at the same time

High volume on its own is not enough. A busy process with unclear rules produces an automation nobody trusts.

It happens often

Enough repetitions that minutes add up to hundreds of hours a year.

The rules are clear

Two people given the same case would do the same thing most of the time.

The data is in a system

The information needed to decide can be read from software, not from someone’s memory or a paper file.

It will still exist next year

Automating a process you are about to change or retire wastes the effort.

The model

How the numbers are worked out

Published in full so you can check it and disagree with it.

Hours a yeartimes per week × minutes each ÷ 60 × 48 working weeks.
Share that can be automatedTypical share for that kind of task, reduced to 70% if rules are mixed or 35% if it needs judgement; to 75% if data is partly in systems or 40% if it is on paper; and scaled by how connectable your systems are.
ValueHours saved × hourly cost, plus 10% or 25% where mistakes are costly.
Ease scoreUp to 40 points for clear rules, 35 for data in a system, 25 for connectable systems.
Approach and costSmall, clear tasks → existing tool (a few days). Ease 70+ → configure (1–3 weeks). 50–69 → integrate (3–6 weeks). Below 50 → custom build (6–12 weeks). Weeks are priced at a typical engineering rate for the currency.
Not worth it yetFewer than 40 hours saved a year, an ease score below 30, or even the optimistic payback is longer than 24 months.

Questions

About this tool

How accurate are the savings figures?

They are planning estimates built from your own volumes and times, adjusted for how consistent the rules are and whether the data sits in a system. Treat them as a way to rank opportunities and size the prize, not as a promise. Measuring one real week of work will tighten them considerably.

Why do some processes say "use an existing tool"?

Because building custom automation for a small process rarely pays back. When a task saves fewer than about 250 hours a year and the rules are clear, a helpdesk macro, a scheduling app or an integration service usually does the job for a fraction of the cost.

What does "not worth automating yet" mean?

Either the time saved is too small, the payback would take more than two years, or the rules and data are too inconsistent for automation to be trusted. The usual fix is to simplify the process or get the data into a system first.

Where do the build cost ranges come from?

They are indicative ranges based on typical effort for each approach: configuring a tool, integrating systems, or building custom automation. Real quotes depend on your systems, security needs and how many exceptions the process has.

Is my data stored?

The calculation runs in your browser. We keep an anonymous record of the answers to improve the model. Nothing identifies you unless you ask for the emailed plan.

Opportunity—View