Begin with observed work
Pick a representative period and count completed cases, failed cases, and cases that returned for rework. Sample active touch time separately from waiting time. Waiting may have business value even when no one is actively working.
Do not use the fastest employee as the baseline or count a rare month as typical. Record a range when volume and handling time vary. Keep the source and date beside every assumption.
Calculate annual manual effort
A simple capacity estimate starts with weekly cases multiplied by active minutes per case and working weeks per year. Divide by 60 to convert minutes to hours.
If 80 cases require 12 active minutes and the process runs for 48 working weeks, the model contains 768 annual manual hours. That is an input, not yet a saving. It ignores exceptions, review, adoption, and maintenance.
Apply a realistic automation percentage
Estimate the share of active work the proposed design could remove or reduce. Keep human review, unresolved exceptions, and work that shifts elsewhere. A workflow that prepares a case may save research time but still require approval.
Use a low, expected, and high scenario rather than one precise percentage. The range should reflect evidence from sample cases and the proposed boundary. If the figure depends on perfect data or complete user adoption, lower it.
| Scenario | Use when | Treatment |
|---|---|---|
| Low | Adoption is uncertain or exceptions are common | Keep more review and rework |
| Expected | Representative samples support the design | Use documented assumptions |
| High | Inputs are stable and failure risk is low | Still include maintenance and review |
Separate capacity, cash, speed, and quality
Reclaimable capacity is not the same as cash saved. Fragmented minutes may let a team respond sooner or absorb growth without adding a role, but they may not reduce payroll. State the intended benefit clearly.
Speed can matter when a lead waits, an order blocks delivery, or an exception delays month-end. Quality can matter when rework, duplicate records, or policy errors are common. Those benefits need their own baseline and measure rather than being converted into a speculative currency figure.
- Capacity: active hours available for other work.
- Cash: spend that is genuinely avoided or removed.
- Speed: reduced queue or cycle time.
- Quality: fewer errors, duplicates, or repeat contacts.
- Control: better traceability, ownership, or policy adherence.
Include the full cost of ownership
A cheap prototype can be an expensive operating system. Model costs over the period the business expects to use the workflow, and include a contingency for known uncertainty. Do not hide third-party charges inside a headline project fee.
- Discovery, process mapping, and implementation
- Workflow, model, database, and integration usage
- Security, legal, and provider review
- Testing, training, documentation, and adoption work
- Monitoring, incident response, and maintenance
- Internal owner time
- Expected rework when source systems change
Use payback and net value carefully
For a cash-based case, estimated payback time is total implementation cost divided by verified monthly cash benefit. If the benefit is mainly capacity or service quality, a cash payback figure may be misleading. Present the operational measure directly.
Net value can compare quantified benefit with implementation and operating cost over a chosen period. Keep each assumption visible and show how the answer changes when volume, adoption, or automation percentage moves.
Turn the estimate into an experiment
The transparent automation estimator on this site calculates manual hours, manual cost, and illustrative reclaimable capacity. It does not include every implementation cost and does not produce a project price. Use it to frame a discussion, then replace inputs with observed data.
- Record the baseline, sample method, range, and owner.
- Define the first workflow boundary and expected human review.
- Choose two or three measures that reflect the real objective.
- Run a controlled pilot and compare equivalent periods.
- Review exceptions, adoption, maintenance effort, and displaced work.
- Update the model and decide whether to expand, revise, or stop.
Frequently asked questions
How do you calculate automation ROI?
Compare verified benefits over a defined period with implementation and operating costs. Keep capacity, cash, speed, and quality separate, and show low, expected, and high assumptions.
What automation percentage should we use?
Use a range based on sample cases, exceptions, human review, and the proposed boundary. A default percentage is only a placeholder until discovery produces evidence.
Should saved employee time be counted as money?
Only when it creates a real financial effect, such as avoided spend or delayed hiring. Otherwise present it as capacity and explain how the team expects to use it.
What costs are commonly missed?
Internal owner time, usage fees, security review, exception handling, monitoring, maintenance, training, and changes to source systems are often omitted from early estimates.
If this topic maps to a real process, document the current volume, exceptions, systems, data sensitivity, owner, and consequence of error before selecting a tool.
Request an automation audit
