⚡ The AI AdvantageGet the book

How much time and money could AI save my business?

Multiply how often each repeated task happens by how long it takes, assume AI takes over only part of it, and subtract the time you spend checking its work and the cost of the tools. Controlled studies found large gains on suitable tasks (40% less time on business writing, 25% faster consulting work, 56% faster on a coding task) and worse results on tasks outside AI's abilities, so start with frequent, clear, low-risk work.

Last updated:

Example tasks are filled in: edit them or add your own from a normal week. Free, no sign-up, nothing leaves your browser.

Your repeated tasks

Share AI takes over starts at a cautious default from "clear steps" (40%, 25% or 10%). Change it if you have tested it.

The full method is in The AI Advantage.
This calculator tells you where the hours are. The AI Advantage shows how to get them: the full Opportunity Audit, choosing an assistant and automation stack, business-grade prompting, sales, marketing and back-office blueprints, guardrails for accuracy and privacy, and a 30-day plan. No coding needed.
See the book

How it works

NumberHow it is worked out
Hours nowTimes a week × minutes each ÷ 60. Ten minutes, forty times a week, is nearly seven hours.
Hours backHours now × share AI takes over, minus your checking time (default 25% of the time saved).
ValueHours back × 52 ÷ 12 × hourly cost, minus monthly tool costs. Payback = set-up hours × hourly cost ÷ monthly value.
Score /15The AI Opportunity Audit: frequency (monthly / weekly / daily), time (under 5 / 5–30 / over 30 min), clear steps, safety of a mistake and customer impact, 1–3 each. 12+ is a strong first candidate; safety 1 goes last.

The default shares (40% for tasks with clear steps, 25% for mostly clear, 10% for judgement calls) are deliberately cautious assumptions, not research findings. Studies measured bigger gains on well-suited tasks, but real businesses lose time to setup, edge cases and checking.

Education, not legal, tax or financial advice. Estimates depend on your inputs; measure real hours before and after. Keep a person in charge of pricing decisions, legal, medical, tax or financial advice, hiring and your most important relationships.

Techniques from the book behind this tool

This tool puts a few of the named techniques from The AI Advantage to work. Each line says how it relates to the tool.

See every technique in The AI Advantage →

Questions

How much time does AI actually save?

On the right tasks, a lot: in a Science study, professionals using ChatGPT for business writing took 40% less time and produced 18% better-rated work; support agents with an AI assistant resolved 14% more issues per hour (34% for the newest agents); consultants did suitable tasks 25.1% faster; developers finished a coding task 55.8% faster. On a task outside AI's abilities, consultants using it were 19% less likely to produce a correct solution.

Which tasks should I automate with AI first?

Score each repeated task 1–3 on frequency, time, how clearly you can write the steps, how cheap a mistake is and whether customers would notice. Totals of 12 or more out of 15 are strong first candidates. Anything where a mistake is expensive or hard to reverse goes last, whatever its score.

What should a small business not automate with AI?

Pricing and discount decisions, anything that is legal, medical, tax or financial advice, hiring and firing decisions, and messages to your most important relationships. AI can draft and summarise for a person; a person decides.

How much do AI tools cost a small business?

Usually a few assistant seats plus an automation platform and some API usage. Set a hard monthly spending limit, use the smallest model that does the job, and compare the bill with the hours saved every month; switch off anything that costs more than it saves.

If AI makes me faster, should I charge clients less?

If you bill by the hour and AI halves your hours, you halve your revenue for the same result. Move towards fixed prices for outcomes or deliverables, so the client gets the speed and you keep the efficiency.

Sources

  1. Noy and Zhang, "Experimental evidence on the productivity effects of generative artificial intelligence", Science 381 (2023). doi.org/10.1126/science.adh2586
  2. Brynjolfsson, Li and Raymond, "Generative AI at Work", Quarterly Journal of Economics 140(2) (2025). doi.org/10.1093/qje/qjae044
  3. Dell'Acqua et al., "Navigating the Jagged Technological Frontier", Harvard Business School Working Paper 24-013 (2023). www.hbs.edu/faculty/Pages/item.aspx?num=64700
  4. Peng, Kalliamvakou, Cihon and Demirer, "The Impact of AI on Developer Productivity: Evidence from GitHub Copilot", arXiv:2302.06590 (2023). arxiv.org/abs/2302.06590

The scoring method, the default task list and the "what not to automate" list come from The AI Advantage (Part 3, the AI Opportunity Audit). The default shares are our cautious assumptions, not study results.