Estimate automation value without inventing an ROI figure

Estimate automation value from observed work and explicit assumptions. Released staff capacity, avoided rework and cash savings are different outcomes and should be reported separately.

Build a baseline that can be repeated

Choose a representative period and count cases using a consistent definition. Record active handling time, waiting, corrections and exceptions separately. If the data is estimated from interviews, say so and identify what a pilot should measure.

Avoid using the fastest successful case as the baseline for a whole process. Include routine variation and the work that happens outside the main system. A workflow that saves entry time but adds a large review queue may move effort rather than remove it.

Calculate capacity before claiming a saving

Illustrative calculation only: 120 cases per month × 6 minutes of reduced handling = 720 minutes, or 12 hours of potential capacity. If new review and exception work takes 4 hours, the net estimated capacity is 8 hours. These figures are fictional and are not a Foxbyte customer result.

Quantity	Question to answer
Case volume	What period and cases does the estimate cover?
Handling reduction	Was it observed using the same definition before and after?
New review effort	What work did the automation introduce?
Realised capacity	Can the team use the released time productively?
Cash effect	Does a real cost change, or is the benefit service capacity?

Include the complete cost

Separate one-off discovery, build, testing and handover from recurring licences, usage, hosting and management. Include the customer’s own preparation and review effort where it is material. State the period used for comparison.

For a Kenyan business paying for foreign-currency services, record the quoted currency and conversion assumption rather than treating an exchange rate as permanent. Confirm applicable tax treatment from the actual quotation. This worksheet does not supply an investment return, a tax conclusion or a market price.

Use a sensitivity range

Recalculate with lower case volume, smaller handling improvement and higher exception effort. Ask whether the project still makes sense under those conditions. The purpose is to expose the assumptions that matter, not to create a precise-looking percentage from uncertain inputs.

Some benefits are operational rather than financial: visible case status, fewer lost requests or more consistent review records. State them as outcomes with acceptance evidence. Do not assign an invented cash value merely to force everything into one ROI total.

Make the pilot answer the uncertain part

• Validate the case volume and active handling baseline.
• Measure review and correction effort as well as successful runs.
• Record failures, exclusions and the sample period.
• Compare the actual recurring costs with the assumptions.
• Decide whether to proceed, revise the scope or stop.

A useful business case survives scrutiny because its inputs are traceable. If a critical assumption is unknown, keep it visible and design the next test to resolve it.