How to Build an AI Business Case

By Paige Gilmore, Founder, NetLift · Published 2026-07-28 · Updated 2026-08-07

Building an AI business case requires measuring the labour value of realised time savings against the total cost of adoption, including licenses, implementation, and rework. Success is defined by current net value and the time it takes to achieve payback on the investment.

A robust AI business case moves beyond hype to focus on deterministic value. To secure CFO approval, you must calculate the labour value of time saved by subtracting the time work takes with AI from the time it took without it, then multiplying that delta by your loaded hourly staff cost.

This approach shifts the conversation from vague productivity gains to net value. By accounting for all costs—including licenses, training, and rework—you can provide a clear payback period and an evidence quality grade that justifies further investment or indicates when to stop.

How do you calculate the labour value of AI?

Value is determined by the time saved on specific work. You calculate this by taking the time the work would take without AI assistance and subtracting the time taken with AI. This resulting figure is then multiplied by the loaded hourly cost of the staff involved.

For example, using a default loaded staff cost of $75 per hour, saving 100 hours of work results in a labour value of $7,500. This provides a concrete financial figure that finance teams can use to assess the impact of AI adoption on the bottom line.

What costs must be included in the business case?

To find the current net value, you must subtract the full cost of AI from the realised labour value. A common mistake is only accounting for license fees. A credible business case must also include usage costs, implementation expenses, initial training, and any time spent on review and rework where it is tracked.

This calculation ensures that the 'net' in net value is accurate. It reflects the actual cost of doing business with AI rather than an optimistic estimate that ignores the hidden overheads of implementation.

How do you prove the data is reliable?

Evidence quality is a critical component of any business case. Numbers should be graded from 'Estimate Only' up to 'Verified' based on the strength of the data behind them. Objective baselines, such as historical data or cohort comparisons, carry more weight than self-estimated savings.

Factors such as sample size, recency of data, and the completeness of cost tracking determine this grade. This transparency allows stakeholders to understand the risk level associated with the projected returns and decide whether to expand, continue, or review a project.

How do you distinguish between realised and future value?

Realised value covers the actual savings and costs incurred to date. Future value is a projection of expected recurring time savings multiplied by expected volume. These two must always be stated separately.

Reporting realised value demonstrates what has already been achieved, while future value outlines the potential for scale. This distinction prevents the business case from becoming speculative and keeps the focus on measurable performance.

NetLift measures work and value rather than individual productivity, avoiding surveillance like keystroke logging or screenshots. By applying a deterministic value model to tracked work, NetLift calculates time saved against objective baselines and assigns an Evidence Quality grade, helping you decide whether to expand, improve, or stop AI initiatives.

Frequently asked questions