Assess a small tool

Use this before committing to automation, a spreadsheet, or a small application. The assessment should include the work required to operate and maintain it.

1. Measure the current task.
Observe several instances. Record frequency, time spent, errors, and consequences. Note differences in difficulty. Distinguish measured values from estimates.

2. Examine existing options.
Check whether an existing tool, template, or process adjustment addresses the difficulty. Identify duplication, access constraints, and required integrations.

3. Define a bounded trial.
State the task the tool would support, its users, inputs, outputs, and expected result. Identify an owner, a cost/time limit, a review date, and a rollback plan.

4. Check the obligations.
Consider data access, accuracy, security, maintenance, support, and approval requirements. Use fictional or otherwise authorized data for early tests. Keep consequential actions under the appropriate human control.

5. Try the smallest useful version.
Test with representative users and cases. Record usage, corrections, exceptions, and support effort alongside task performance. Preserve a working fallback during the trial.

6. Make the continuation decision.
Compare the observed benefit with setup, training, maintenance, and support costs. Check whether people actually use the tool. Continue, revise, defer, or stop, with reasons recorded.

WORKSHEET

Task, users, and frequency:


Observed time / error / consequence baseline:


Measured values and estimates:


Existing options considered:


Inputs, outputs, access, and owner:


Trial scope, cost limit, and review date:


Expected result and stop conditions:


Fallback and maintenance responsibilities:


Usage, corrections, and support observed:


Continuation decision and reason:


Check adoption, correctness, time saved, support demand, and who carries the remaining work. Keep estimated benefits labeled until measured. Stop or revise when the tool causes errors, needs more support than available, or no longer addresses the task.

REASONING

The recovery period depends on actual correction and support time, how often the report is used, and how long its format remains useful. Record those assumptions during the trial. Include the effort required for someone else to operate and maintain the tool.

Try the reasoning on another instance

Ask a colleague to apply the worksheet to another safe instance and explain their decision. Record what they found unclear and when they would ask for help.

Explanation being tested:


Other possible explanation:


Observation that would change the decision:


What another person needed to use this method: