Measure Before and After

Why this matters

If you do not measure, you are just rearranging steps and hoping. The whole point of the smallest useful change is that its effect can be seen. Without the numbers, the only opinions that survive are the loudest ones — and the person who owns the changed step will defend it forever, because nobody can prove otherwise. Measurement is how you keep the argument honest.

What you will be able to decide after this

Whether the change worked, on the numbers you named before you made it.
Whether the change is responsible for the improvement — or something else is.
Whether to keep, extend, or roll back the change.

Core lesson

Define a small set of numbers before you implement:

Cycle time or lead time — from real trigger to done, as the map defines it.
Error or rework rate — the loops from Module 1, counted.
Volume handled — how much work the workflow carried, per week.
Time spent on verification or exception handling — the cost of the judgment steps, which is the cost a wrong change inflates silently.
Any stakeholder pain measure that matters — the complaint that actually shows up in your meetings, whatever it is.

Four or five numbers is a full set. More than that and the sheet becomes decoration.

Collect the baseline — you should already have most of it from Module 1. If a number was missing there, collect it now, before the change. The baseline must be collected before implementation. There is no legitimate "we forgot the baseline, let's estimate it later."

After the change has run long enough to be real, collect the same numbers again. "Long enough to be real" is a cycle length judgment: a workflow that runs weekly needs a few weeks; a monthly one needs at least two cycles. If the numbers are collected after two days of enthusiasm, they measure attention, not the change.

The before/after sheet:

Baseline — the Module 1 numbers.
Target — the number you committed to in the Module 3 brief.
Actuals — the same numbers after the change has run.
Confounding factors — what else moved while the change ran. Lower volume, different people, a deadline that focused everyone, the season. If you cannot name any, you have not looked.

Worked example

The renewal workflow's sheet: baseline specialist wait 14 hours, target under 4, actual 3.2 after three weeks. Baseline first-pass acceptance 78%, target 90%, actual 88%. Rework rate flat at 5%. Volume flat — the confounding check matters here: the team's volume did not drop, the same specialists handled the cases, and no deadline or process holiday coincided with the window. The improvement tracks the change: the wait collapsed exactly where the draft step and the twice-daily review sit. If volume had halved in the same weeks, the wait would have dropped regardless of the change — the sheet would show that, and the honest conclusion would be "measure again when volume returns."

Practice

Write the five numbers for your workflow, the target for each from your Module 3 brief, and a list of plausible confounding factors for the measurement window. One sentence each.

Apply — produce the artifact

Create the before/after measurement sheet for the change you designed in Module 3. Include the baseline numbers, the target, and the actuals once available. Note any confounding factors. If a baseline number is missing, collect it before the change runs.

Verify

If the after numbers are better, check whether the improvement came from the change you made or from something else — lower volume, different people, temporary attention. If you cannot tell, the measurement is incomplete. Write the check down.

Sources

This module is original practice guidance based on the authoring standard and does not depend on a specific external factual claim. Editorial review is still required.