Understanding change and making its case · Lesson 4

Defining a measurable future state

4 min read

Lesson objectives

  • Frame a measurable future state and justify the decision using case evidence.
Reading tools

A change initiative may finish launching a new form, only for most staff to stop using it two weeks later. Do not confuse **delivering the mechanism** with **achieving the result**. A future state describes how people will work, what the user will experience, what outcome is expected and how we will know it lasts. “We now have one system” describes a technical asset. “Requests are entered once and reach the reviewer with complete information” describes observable work. We need measures of behavior and impact, plus guardrails that reveal harm hidden by a headline number.

Organize a small measure register into four kinds. An **implementation measure** shows that the mechanism is ready, such as a working form and trained team. An **early adoption measure** shows correct use, such as the share of eligible requests that follow the new handoff without an unjustified exception. An **outcome measure** shows a later effect, such as fewer returns caused by missing data. A **guardrail** protects another concern, such as customer waiting time or errors affecting a vulnerable group. Training attendance is activity, not demonstrated ability; a system login does not establish data quality. Select indicators connected to the causal hypothesis rather than filling a dashboard until it obscures the decision.

For every indicator, write an operational definition: numerator, denominator, period, data source, owner and review frequency. “Adoption is 80%” might mean 80% of employees, requests or sites; those are different claims. If the dashboard includes only easy cases that use the system and silently removes exceptions, success will be overstated. Describe legitimate exceptions, record them separately and show what happens to the rate when they are included. A measure another person can recompute keeps the discussion on the work instead of on trust in a dashboard.

Set a **decision point** before the pilot begins: when results are reviewed, what condition triggers expansion, modification or stopping, and who has authority. Choose neither an impossible target nor a trivial one. If the baseline return rate is 20%, an organization might set a local pilot aim of 15% after a month while keeping waiting time within a limit informed by its baseline. Those numbers are fictional choices, not universal benchmarks. Inspect distribution as well: the average might improve because a large team does better while a small team delivering a sensitive service does worse. Review actual work samples alongside the chart.

Measures can change behavior. Rewarding closure speed alone may encourage staff to close incomplete cases or shift work to reviewers. Rewarding zero returns alone may encourage avoidance of complex cases. That is why a quality guardrail and qualitative feedback from staff and users matter. If the rate improves while new complaints appear, do not settle the conflict by choosing either “data” or “anecdote.” Check how each was collected and trace the workflow that produced both, then revise the intervention or the measure.

Lesson summary

Frame a measurable future state with stated assumptions and verification.