The visible price is only part of the operating picture
An AI subscription may be easy to understand, while the effort required to make it useful is less visible. A business may need to prepare information, connect systems, train employees, and review outputs. Including that work early gives the owner a clearer basis for deciding whether the project is practical.
Think in terms of the completed business task. The cost of producing a draft matters less if checking it takes longer than the old process. Conversely, a system that requires some setup may be useful if it removes repeated work across a frequent, stable workflow.
Separate setup effort from recurring work
List the activities needed to get started: discovering the process, preparing suitable information, configuring access, connecting systems, testing representative examples, and teaching the new routine. Note who will do each activity and how it affects their existing work. Internal employee time is still a real demand on the business.
Then list the work that continues after launch. This can include subscriptions, usage charges, monitoring, human review, updating source material, fixing connections, and answering employee questions. The GOV.UK technology guidance includes total ownership cost and flexibility among the considerations when choosing technology.
Model ordinary use and difficult cases
Usage can change with task length, the number of steps, and the amount of review needed. Compare a typical period with a busier period. Also consider exceptions that trigger repeated attempts or require specialist attention. Make assumptions visible so the team can replace them with observed information during the pilot.
In a hypothetical customer-service workflow, an assistant might prepare routine answers quickly but require extensive checks for unusual requests. The business should estimate the mix of both types. Measuring only the easy cases could create an unrealistic picture of the resources needed for everyday operation.
Describe benefits in terms the business can realize
Time saved is useful capacity, but it does not automatically reduce expenditure. Decide how the business would use that capacity: respond sooner, handle more work with the same team, reduce overtime, or give employees time for follow-up. These are different outcomes and need different evidence.
Avoid counting the same benefit twice. Faster preparation and more available staff time may describe the same improvement. Track quality and customer experience alongside speed so the business can see whether the new process reduces effort overall or simply transfers it to corrections, complaints, or another department.
Include change, exit, and ongoing decisions
Ask what happens when the business changes a service, replaces a system, or wants to move its information. Check the effort required to export records, revise the workflow, and retire the automation. A solution that can be adjusted in small steps may be easier to manage than one whose value depends on a long, inflexible commitment.
Use the pilot to replace assumptions with real observations and set a review date after launch. A practical cost picture does not require false precision. It requires a complete enough view of setup, operation, and change to help the business choose a sustainable next step.
Estimate the effort of delivering the complete task, then validate recurring usage, review time, and support needs during a realistic pilot.
Sources & further reading
Primary references for the concepts discussed here. Any business examples in this article are illustrative; results depend on your process, tools, and implementation.