AI Boss Tools | Calculator • Back to the Hub
Most AI budgets price the software and forget what the change costs, which is where they overrun. This calculator works backward from the result: name the annual value you want in savings, new revenue, and retention, and it returns the investment that implies, split into direct AI cost, change management, and a risk provision. You also get the yearly run cost, payback, and return over two, three, or five years.
Estimated First-Year Investment
Enter a value target below, then press equals.
The value you want, per year (USD)
Your context
Rough figures are fine. The calculator updates as you type, and nothing you enter leaves your browser.
Enter your target value above and the breakdown will appear here: the three cost buckets, the full line items, your time to value, and your payback period.
Licences, models, data work, and integration. The part a vendor quote usually covers.
Training, process redesign, communication, and the productivity dip while people adjust. The part most budgets miss.
A buffer for governance, security, oversight, and rework, sized to your sector and data sensitivity.
| Line item | Estimate |
|---|
A benefit-to-cost ratio turns your target value into a delivery investment, then splits it on the BCG 10-20-70 rule: about a third to technology and data, the rest to people and process. Change management is floored against the number of people affected, an AI risk provision is sized to your sector and data sensitivity, and the cash flow runs month by month, so value starts at go-live and the expected case keeps the share of target most programmes actually capture. Everything runs in your browser and nothing you type is sent anywhere.
Sources: BCG 10-20-70 (about 70% of the value is people and process); IDC and Microsoft (about US$3.70 returned per dollar invested in generative AI); IBM, Cost of a Data Breach (about US$670,000 in added cost where shadow AI is involved); MIT, State of AI in Business 2025 (about 95% of enterprise AI pilots show no measurable profit); Gartner (at least 30% of GenAI projects dropped after proof of concept).