Decision Factory / Brief 001
Should we buy
the support automation?
The proposal looks attractive at first glance. The answer changes when we model the work that can actually be automated.
Commentary on the original case · 20% eligible / 60% resolved / 2 min review
What the first recommendation gets right.
Support is a plausible place to examine automation. At 120,000 tickets, eight minutes each, and $45 per hour, the baseline handling capacity is worth $720,000 annually.
But that number describes the entire workload. It is not the value recoverable by the proposed product.
What changes the call.
Under the original assumptions, only 20% of tickets are eligible, 60% of that subset is resolved, and each eligible ticket creates two minutes of additional review. Those constraints reduce the recoverable capacity value to $50,400.
The proposal costs $164,000 in its first year before additional internal implementation work. Under these assumptions, the recommendation is to decline the annual rollout.
The alternatives belong in the decision.
Keeping the current workflow is the baseline. Buying the proposed annual rollout adds $164,000 of first-year cost that these assumptions do not justify. A narrower purchase may deserve comparison if its full cost and recoverable value differ materially. Building internally also has implementation, maintenance, and review costs; it is not a free alternative.
Before commissioning more work, the support owner should identify the eligible cohort, finance should establish how recovered capacity would create value, and engineering should identify integration and operating obligations. That evidence determines whether an alternative merits a scoped evaluation.
What evidence could change the recommendation?
- Eligibility: a representative sample showing which tickets the product can safely handle.
- Resolution and review: observed outcomes that include exceptions, escalations, and quality checks.
- Realizable benefit: a named plan to avoid spending or use the recovered capacity for measurable value.
- Commercial terms: a materially different price, scope, or operating cost.
A paid pilot is not automatically the right next step. First determine whether affordable evidence could plausibly change the economics enough to justify one.
This is a worked fictional example, not a customer case, a forecast, or an offer to perform an implementation. Changing the sliders updates the model and recommendation above; this narrative describes the original case.