Methodology
Evidence before enthusiasm.
We evaluate AI through the workflow it must complete—not the quality of its demo.
1. Define the job
Every test begins with a specific business outcome, required inputs, acceptance criteria, and the human effort currently required.
2. Record the full cost
We include subscription or API cost, setup time, integration work, supervision, retries, and the cost of failure.
3. Run repeatable trials
Where outputs vary, we use multiple trials and preserve prompts, settings, source material, dates, and model versions.
4. Separate capability from reliability
A tool that succeeds once may be capable. A tool that succeeds consistently, safely, and economically may be deployable.
5. Publish limitations
Recommendations include source links, assumptions, failure modes, and a last-verified date. Material vendor relationships will be disclosed.
How we calculate ROI
The AI ROI calculator values recovered time at its loaded cost and compares it with the monthly software bill. It uses four inputs: employees affected, hours saved per employee each week, loaded hourly cost, and monthly AI software cost.
- Weekly gross savings = employees × weekly hours saved × loaded hourly cost
- Monthly gross savings = weekly gross savings × 4.33 weeks
- Monthly net savings = monthly gross savings − monthly software cost
- Annual net savings = monthly net savings × 12
- ROI on software cost = monthly net savings ÷ monthly software cost × 100
- Payback period = monthly software cost ÷ weekly gross savings × 7 days
The results are rounded to one decimal place. When net savings are zero or negative, the calculator shows no payback. When the software cost is zero and time savings are positive, it reports ROI as immediate.
The model counts only hard time savings against a recurring software cost. It does not include one-time implementation, integration, training or change-management costs. It does not count revenue or quality gains either, and it assumes recovered time can be redeployed. In line with step 2 above, add those costs to your business case and discount any hours that can't be reassigned. Validate hours saved with a measured pilot before scaling.