How to measure automation value without inventing the business case
A credible case links operational baselines, adoption and quality. It does not multiply theoretical minutes by salary and call the result cash.
Value realisationAutomation value should be reported as a bridge from baseline performance to observed operational change, with capacity, quality and cash effects kept separate.
01Time released is capacity, not automatically cash.
02Quality and demand effects may matter more than labour savings.
03Benefits need an operational owner and an agreed route to value.
Begin with the counterfactual
A business case needs a clear description of what would happen without the change. Record volume, cycle time, error, rework and service performance over a representative period.
Avoid baselines built from one unusually difficult week or staff estimates alone. Where data is weak, state the uncertainty and improve measurement during discovery.
Separate benefit types
Minutes released create capacity. They become cash only if cost is removed, avoided or redirected to work that produces measurable value. Keep these cases separate.
Quality improvement may reduce rework, complaints or risk. Service improvement may increase conversion or retention. Each benefit requires its own causal logic and evidence.
Measure adoption and exceptions
A technically successful workflow produces little value if people work around it. Track eligible volume, actual use, completion and the reasons users revert to the prior process.
Exception demand is equally important. A system that automates routine work but doubles complex rework may have negative total value.
Create a benefits cadence
Assign an operational owner to each material benefit. Review the evidence at defined intervals and retire measures that do not affect decisions.
A transparent case can still support investment when uncertainty is high. It should show ranges, assumptions and the evidence required to narrow them.
What the wider evidence says
Findings are paraphrased from the linked original publications. Their scope and populations differ, so they inform the thesis rather than prove a universal outcome.
A field study of 5,172 support agents found higher issues resolved per hour with AI assistance
Value can be material in a well-matched workflow, but the measured outcome is specific to the task and operating environment.
Source: Quarterly Journal of Economics, Generative AI at Work, 2025A different randomised study found a slowdown for experienced developers on familiar repositories
A credible business case must test the target workflow rather than import a productivity percentage from another context.
Source: METR, Experienced Developer Productivity Study, 2025Google DORA connects returns to the quality of the organisational system
Benefits measurement should include adoption, process quality and the capabilities surrounding the tool.
Source: Google DORA Report 2025Most UK adopters report productivity improvement while most report no revenue change
Business cases should distinguish operating performance from realised financial value and make the conversion mechanism explicit.
Source: DSIT, AI Adoption Research, 2026A controlled route from thesis to operating evidence
Each work package ends with an explicit decision and a tangible output. The sequence keeps delivery connected to operating evidence.
Build the counterfactual
- Decision
- What would performance look like without the change?
- Work
- Select a representative period and record volume, effort, delay, quality, service and risk.
- Output
- A baseline with known uncertainty and data limitations.
Write the causal chain
- Decision
- How should the intervention create each benefit?
- Work
- Separate capacity, cost, revenue, quality, service and risk mechanisms.
- Output
- A benefit hypothesis with owner and disconfirming evidence.
Observe full operating cost
- Decision
- What effort and cost does the new process add?
- Work
- Measure review, exceptions, workarounds, support, suppliers and change effort.
- Output
- A net operating view rather than gross time saving.
Govern realisation
- Decision
- Should the business expand, adjust or stop investment?
- Work
- Review evidence, confidence, adoption and conversion of released capacity at a fixed cadence.
- Output
- A benefits ledger and explicit management decision.
The failure modes leadership should watch before scale
Risks become manageable when the early signal, control and accountable owner are agreed before release.
A scorecard that connects activity to management action
Measures are useful only when their definition is stable and their movement changes a management decision.
Released capacity is not the same as realised cash
Illustrative bridge from gross time saving to evidenced value.
A benefits ledger keeps assumptions and evidence together
01baseline = metrics.window(before)02observed = metrics.window(after)03delta = adjust(observed - baseline, demand)04value = benefits.classify(delta)05ledger.record(value, owner, confidence)