Understanding change and making its case · Lesson 1

Separating a symptom from the change problem

4 min read

Lesson objectives

  • Separate the presenting symptom from a testable change problem and justify the decision using case evidence.
Reading tools

An operations manager receives a message: “Rework has risen; we need a new system immediately.” The observation deserves attention, but the proposed system is a decision made before diagnosis. Rework is a **symptom**. It might result from incomplete intake, a changed definition of completion, uneven workloads, or a software defect. A useful change problem states what is happening, to whom, over which period, by which measure, and what remains unknown. That description gives the team a basis for choosing an intervention instead of committing to the most visible solution.

Start with the unit of measurement. If 28 of 100 cases closed during a month were returned for correction, the rework rate is 28%. Yet “of 100” needs a definition. Is a case counted once at its first closure or at every closure attempt? Are cancelled cases included? If last month’s denominator was all cases received, the two rates are not directly comparable even if both divisions are correct. Record numerator, denominator, period and data source together. A percentage alone hides choices that can reverse an apparent trend.

Next, inspect the distribution. An organization-wide average can suggest a general problem even when rework is concentrated in one team or one case type. Break the records down by task, site and period, but do not confuse association with cause. A rise after a tool launch does not prove that the tool caused it. The definition of “returned” may have changed, or incoming requests may have become more complex. Ask staff to describe the work they actually perform, inspect sample records and compare observed handoffs with the intended procedure. Different evidence sources are useful precisely because they may disagree.

A defensible problem statement for this case is: “During the test month, 28 of 100 cases closed for the first time by the intake team were returned. We need to test whether missing handoff fields, the closure rule or the case mix explains the increase before choosing an intervention.” The statement does not pretend to know the cause; it names a test. Suppose 18 returned records lack a required field and observation shows that conflicting instructions cause the omission. Redesigning the handoff and giving focused practice may be more useful than buying software. If instructions are clear and fields are complete but the system drops data during transfer, the decision changes. Diagnosis matters because a material change in evidence can change the appropriate action.

Waiting for evidence does not mean letting harm continue. A temporary containment step, such as checking ten cases each day before closure, can have an owner and an explicit end date. Keep containment separate from the permanent remedy. A manual check that lowers rework does not establish that the underlying cause is gone. Monitor its effect on service time too: rework might fall because closure slowed down or because staff avoided difficult cases. A balanced view includes rework rate, processing time and case mix rather than a single headline number.

Lesson summary

Separate the presenting symptom from a testable change problem with stated assumptions and verification.