The Owner Bottleneck · Chapter 3

Measure before you automate

A changed workflow is not automatically an improved workflow. Establish a small baseline, test one contained operating change, and compare the result without creating measurement bureaucracy.

Build the baseline

← Chapter 2: How to Choose the First Workflow to Improve

A faster process can still be worse

Speed can improve while rework increases. A new form can make handoffs more complete while creating unnecessary entry. Automatic notifications can reduce owner questions while overwhelming employees.

Measure before automating so the business can recognize improvement rather than merely recognize change.

A baseline measures what the workflow currently requires. It is not a score of whether people are working hard enough.

Measure the movement of work

Performance includes speed, reliability, quality, rework, customer impact, cost, and the ability to complete a trustworthy handoff.

Time

Elapsed and waiting time

Separate how long the full cycle takes from how long it cannot advance.

Reliability

Handoffs and rework

Observe whether information arrives complete and whether steps must be repeated or corrected.

Dependence

Owner interventions

Count routine questions and reconnections while labeling genuine exceptions and judgment calls.

A small baseline is enough to begin learning

Use comparable operating cycles, not a calendar period selected only because reports normally use weeks or months.

1 · Define

Preserve the boundary

Use the same starting event and completed outcome before and after.

2 · Observe

Choose two to four measures

Define each unit and record comparable cycles and exceptions.

3 · Test

Change one meaningful thing

Write the hypothesis and identify what must not become worse.

4 · Compare

Repeat the same measurement

Keep, adjust, remove, or clarify the change based on the direction.

Before-and-After Baseline Worksheet

Measure enough to support the next decision

Use the same boundary, definitions, and units before and after one contained operating change.

1. Workflow and desired result

Selected workflow: From ____________________ to ____________________

Current consequence: This workflow repeatedly creates ____________________________

Desired operating result: We want ________________________________________

One workflow cycle is: ____________________________________________

Comparable before and after cycles: ________________________________

2. Select and define two to four measures

Possible measurePractical definitionUse it?Exact definition and unit
Total elapsed timeTime from starting event to completed outcome
Waiting timeTime the workflow is unable to advance
Owner interventionsRoutine questions, approvals, status explanations, or reconnections
Rework or correctionsSteps repeated or corrected because information or execution was incomplete
Repeated entry or chasingInformation entered again or requested after it should have been available
Complete handoffsCycles reaching the next person or system with required information
Customer consequenceObservable missed response, update, commitment, or customer effect
Financial handoff readinessCycles reaching the appropriate financial process complete and on time

Keep the financial boundary clear

Operational teams may observe whether a handoff is complete and timely. Accounting treatment and financial judgment remain with the appropriate qualified professionals.

3. Before baseline log

Replace the generic measure headings with the exact units selected above. Add rows when the workflow needs more cycles.

Cycle/dateMeasure 1Measure 2Measure 3Measure 4Exception or context
 
 
 
 
 
 

4. Summarize the before baseline

Choose the summary that fits the measure: a total, count of affected cycles, typical result, range, or proportion.

Selected measureBefore resultWhat it appears to showException or limitation
 
 
 
 

5. Define one change

Change to test: We will ____________________________________________

Hypothesis: If we ____________________, then ____________________ should improve without making ____________________ worse.

What stays unchanged: ___________________________________________

Test length or cycles: ____________________________________________

6. After-test log

Use the same workflow boundary, measures, definitions, and units.

Cycle/dateMeasure 1Measure 2Measure 3Measure 4Exception or context
 
 
 
 
 
 

7. Compare before and after

Selected measureBeforeAfterDirectionTradeoff, exception, or limitation
 Improved / unchanged / worse / unclear
 
 
 

8. Make one evidence-based next decision

  • ☐ Keep the change and observe a few more cycles.
  • ☐ Adjust the change and repeat the comparison.
  • ☐ Remove the change because it added effort, risk, delay, or uncertainty.
  • ☐ Improve the definition or collect more comparable cycles.
  • ☐ Escalate an exception or risk to the appropriate responsible professional or authority.

Hypothetical example

Completed visit to invoice readiness

These numbers demonstrate the method. They are not an industry benchmark or proof that the change caused every difference.

Before · 10 visits

Current baseline

  • Total waiting: 31 hours
  • Owner follow-ups: 6
  • Complete handoffs: 4 of 10

After · 10 visits

Contained operating test

  • Total waiting: 9 hours
  • Owner follow-ups: 1
  • Complete handoffs: 8 of 10

The result supports a decision to continue observing or adjust the test. It does not manufacture certainty from a small sample.

Keep the measurement proportional

  • Measure the workflow, not employee worth or effort.
  • Use small samples for operational learning, not statistical proof.
  • Record unusual cycles and limitations instead of hiding them.
  • Do not collect data whose administrative cost exceeds the decision it supports.
  • Remember that these observations are not automatically KPIs.

Technology in its proper role

Technology can make timestamps reliable, require critical information, connect handoffs, reduce repeated entry, and preserve comparable records. It can also generate large quantities of activity data that nobody needs.

Start with the decision the measurement must support. Automate collection only when it reduces effort and improves reliability without creating surveillance or administrative burden.


Next chapter

The Difference Between a Metric and a KPI

Decide which observation deserves an objective, threshold, owner, and predefined action.

← Explore all insights

Measure enough to learn—not enough to create another job

A small baseline gives the business evidence for the next decision. Keep the measures that clarify action, and stop collecting those that only create activity.

Talk through one workflow