Test the economics with your own assumptions
This calculator models labour capacity, rework reduction, optional commercial contribution, recurring costs and implementation spend. It is deliberately transparent: every output can be traced to an input below.
Operational automation ROI calculator
Indicative scenario—replace every default with observed data.
Hover, focus or tap an info icon for guidance.
- Labour capacity value / monthCapacity hours multiplied by loaded hourly cost. Only treat this as cash value when the capacity avoids spend or is deliberately redeployed.
- £5,200
- Rework reduction value / monthEstimated correction hours avoided each month multiplied by loaded hourly cost.
- £480
- Gross benefit / monthMonthly labour capacity value, rework reduction value and optional commercial contribution before running costs.
- £5,680
- Year-one net benefitTwelve months of gross benefit minus implementation cost and twelve months of running cost.
- £20,160
Treat the result as a scenario, not a forecast. Use observed workflow data, validate commercial contribution with finance and test at least three cases before approving investment.
For AI email order processing, start with monthly order volume and the minutes spent reading emails, checking attachments and rekeying details into your ERP. Include the human review time that remains after automation and the cost of correcting order-entry errors.
Build value from four separate buckets
Automation business cases become unreliable when every benefit is collapsed into “time saved”. Separate the mechanisms so finance and operations can challenge each assumption independently.
Capacity returned
Handling time removed from repetitive work. This becomes financial value only when capacity avoids spend or is redeployed deliberately.
Quality & rework
Time and direct cost avoided when validation, consistency and exception handling reduce corrections or failures.
Throughput & contribution
Additional contribution from responding sooner, processing more demand or preventing work from ageing out of the queue.
Cost avoidance
Future hires, overtime, outsourcing or system spend the operation no longer needs as volume grows.
Use a formula people can inspect
The calculator uses a simple first-year model. It is appropriate for screening a contained operational workflow before a more detailed investment appraisal.
“Loaded hourly cost” should reflect the cost your organisation uses for decision-making, which may include salary, employer costs, benefits, facilities and management overhead. Do not substitute a revenue rate for a labour cost.
Include the costs that disappear from vendor demos
Model or software fees are only one part of the cost. Operational automation changes processes, permissions, responsibilities and exception handling, all of which need to be designed and maintained.
- Process discovery, data mapping and workflow design
- Integration build, testing and system access
- Exception handling and retained human review
- Model, API, infrastructure and observability costs
- Training, change management and operating documentation
- Ongoing evaluation, maintenance and control improvement
Build the baseline before modelling the future
Use a representative period long enough to include routine work, peaks and exceptions. Four to six weeks is often a useful starting point, but seasonal operations may need a longer view.
| Measure | How to observe it | Common mistake |
|---|---|---|
| Volume | Count completed and abandoned items by type | Using average volume without peaks or mix |
| Handling time | Sample active work separately from queue time | Counting elapsed time as labour time |
| Rework | Record cause, correction time and downstream impact | Treating hidden corrections as normal work |
| Service outcome | Track response, completion and exception age | Measuring task speed but not customer value |
| Cost | Agree loaded rates and future spend with finance | Assuming returned capacity equals cash |
A positive ROI is a prompt for scrutiny, not approval
Run a conservative, expected and upside case. Vary the assumptions that drive the result most: eligible volume, time genuinely removed, adoption, error reduction, retained review and recurring operating cost.
Conservative
Lower eligible volume and benefit; higher delivery time, review effort and running cost.
Expected
The evidence-backed case used for planning, with explicit owners for every assumption.
Upside
Additional value that is plausible but should not be required for the core decision to work.
Replace forecast values with live deployment data as soon as possible. The business case should become an operating dashboard: throughput, touch time, exceptions, quality, cost and the commercial outcome the workflow was meant to improve.
HM Treasury’s current appraisal guidance also recommends sensitivity analysis and explicit adjustment for optimism bias. Although most commercial automation projects do not need a public-sector Green Book appraisal, the discipline is useful: assume costs may be higher, benefits lower and delivery slower than the first estimate.
See the official HM Treasury Green Book 2026 for the underlying principles on costs, benefits, uncertainty and optimism bias.
Before approving the first deployment
- Is the baseline observed, representative and split by meaningful work type?
- Does each benefit have a named owner and a measurement method?
- Have capacity, cash saving and commercial contribution been kept separate?
- Are implementation, retained review and steady-state operating costs included?
- Does the conservative case remain acceptable?
- What result would cause the project to pause rather than expand?
- When will forecast assumptions be replaced with live data?
- Is the workflow valuable enough to matter but contained enough to measure?