Eroding Goals

Thinking Tools3.3 Mapping the System Around the ProblemAlso known as Drifting goals

One of the systems archetypes in Peter M. Senge's The Fifth Discipline (1990), later set out by Daniel H. Kim and Virginia Anderson (1998) and John D. Sterman (2000).

Describes how a gap between target and performance closes by lowering the target, with performance unchanged, so standards slip a little at a time.

The on-time gap closes, but by lowering the target

A gap between target and performance can close two ways: fix the work (B1, slow) or lower the bar (B2, quick). At load the pressure to lower the target wins: Play, and the gap all but closes while performance stays flat, because the target came down to meet it. Set the pressure to zero (a target anchored outside the team) and raise the fixing effort to see the gap close the other way. Illustrative.

Select a loop to trace it, or a box to read it.B1, Fix the work. A gap prompts changes to the work, which raise performance and close the gap. It works, but slowly. One negative (−) link: an odd count, so the loop balances.B2, Lower the bar. A gap prompts pressure to lower the target, which closes the gap at once and changes nothing about the work. One negative (−) link: an odd count, so the loop balances.On-time target. The share of deliveries promised on time, set at 95% in January.Gap. Target minus performance: ten points in January.Pressure to lower it. Each month the gap shows on the dashboard, someone argues the target was unrealistic.On-time performance. The share of deliveries actually on time: 85% in January.Fixing the work. Route planning, loading, and staffing changes. They take months to show.
20%
5%

Gap, points

0: 10.01: 7.52: 5.63: 4.24: 3.25: 2.46: 1.87: 1.38: 1.09: 0.810: 0.611: 0.412: 0.30.02.55.07.510.001234567891011120: 10.01: 7.52: 5.63: 4.24: 3.25: 2.46: 1.87: 1.38: 1.09: 0.810: 0.611: 0.412: 0.30.02.55.07.510.00123456789101112

Performance, % on time

0: 85.01: 85.52: 85.93: 86.24: 86.45: 86.56: 86.67: 86.78: 86.89: 86.810: 86.911: 86.912: 86.90.025.050.075.0100.001234567891011120: 85.01: 85.52: 85.93: 86.24: 86.45: 86.56: 86.67: 86.78: 86.89: 86.810: 86.911: 86.912: 86.90.025.050.075.0100.00123456789101112

Month 12 of 12

Gap, points
0.3
Performance, % on time
86.9
Target, % on time
87.3

B1, Fix the work
A gap prompts changes to the work, which raise performance and close the gap. It works, but slowly. One negative (−) link: an odd count, so the loop balances.
B2, Lower the bar
A gap prompts pressure to lower the target, which closes the gap at once and changes nothing about the work. One negative (−) link: an odd count, so the loop balances.
On-time target
The share of deliveries promised on time, set at 95% in January.
Gap
Target minus performance: ten points in January.
Pressure to lower it
Each month the gap shows on the dashboard, someone argues the target was unrealistic.
On-time performance
The share of deliveries actually on time: 85% in January.
Fixing the work
Route planning, loading, and staffing changes. They take months to show.

An example

A regional distributor set a target of 95% of deliveries on time for the year. In January it ran 85%. The operations dashboard showed the gap in red every month, and every month someone in the review made a reasonable case: the winter routes were harder, a new depot was still settling in, the target had been set before the fuel surcharge. By June the target on the dashboard read 89%, the gap was nearly gone, and the review had stopped discussing it.

A new finance director asked a plain question: had the gap closed because performance rose or because the target fell? Performance had moved from 85% to 87%. Six of the eight points of progress were the target coming down. The company reset the target to 95%, fixed it for the year with any change needing the board's sign-off, and reported it beside the industry benchmark. The routing and loading changes that had been planned and never funded were funded that quarter.

Why it works

A gap between a goal and performance creates tension, and the tension can be released two ways. Fixing the work closes the gap by raising performance, but it costs money and effort and takes months to show. Lowering the goal closes the gap at once and costs nothing, and each step down sounds reasonable on its own. Under pressure, the quick loop wins.

The drift is hard to see because it never looks like a decision. No one votes to accept worse results; the target slides a point at a time, each with a good reason, and the dashboard turns green. The archetype names the pattern and gives one test for any closed gap: which line moved, the result or the target. Anchoring the goal outside the team (a benchmark, a fixed date, a board's sign-off) cuts the quick loop and leaves only the slow one.

The thinking behind it

Peter M. Senge Peter M. Senge, a senior lecturer at MIT Sloan who brought system dynamics to managers, named eroding goals among the system archetypes in the appendix of The Fifth Discipline (1990). The archetypes are recurring feedback structures that show up across very different organizations; this one, two balancing loops with one acting on the goal itself, explains why standards slip without anyone choosing to lower them.

Daniel H. Kim Daniel H. Kim and Virginia Anderson set the archetype out as a practical diagnostic template in Systems Archetype Basics: From Story to Structure (1998), a Pegasus Communications workbook that teaches each archetype through worked stories and causal loop diagrams. Kim's guides and his articles in The Systems Thinker turned archetypes like this one into tools a team can use on its own situation, starting from the story and drawing the loops.

John D. Sterman John D. Sterman, director of the MIT System Dynamics Group, showed with the beer distribution game that managers misread feedback and delays, ordering in ways that create wild swings in inventory (Management Science, 1989). His Business Dynamics (2000), the standard textbook of system dynamics, covers the stocks, flows, feedback loops, and delays of a system and the archetypal structures they produce, this one among them.

How to run it

  1. List the gaps that have closed. Every target that moved from red to green in the last year.
  2. Ask what closed each one. Plot the target and the performance on the same chart and see which line moved.
  3. Name the pattern where the goal moved. Say it aloud in the review: the gap closed because the target fell.
  4. Reset the goal to what the work requires. Customers, regulators, or the strategy set the standard, not last quarter's results.
  5. Anchor the goal outside the team. Tie it to a benchmark, freeze it for the year, or require sign-off from someone not measured on it.
  6. Fund the slow loop. Close the gap through the changes to the work that the lowered target let everyone postpone.
  7. Allow a reset only in writing. A target can change when the facts do, with the reason recorded and visible.

Where it goes wrong

Every target gets frozen. A goal that was wrong to begin with is defended as a standard, and the team learns that targets are fiction. Allow changes, but make each one a recorded decision with a reason, not a monthly drift.

The drift is in the definition, not the number. The target stays at 95%, but on time quietly comes to mean within a day of the promised window. Check the definitions behind any metric that has improved without a change in the work.

The review blames the people who argued for the lower target. Each argument was reasonable; the structure turned them into a slide. Fix the structure, not the people.

When to use something else

  • Goodhart-Aware Metric Selection. When the target holds still but people game the measure instead of improving the work.
  • Shifting the Burden. When a quick fix relieves the symptom and the real fix is neglected, but the goal itself is not moving.
  • Control Chart. When the question is whether a change in performance is real or just ordinary variation.

Where it sits in SOLVE eX

Go to the source

  • The Fifth Discipline, Appendix 2, systems archetypes, 1990
  • Systems Archetype Basics, Pegasus Communications, 1998
  • Business Dynamics: Systems Thinking and Modeling for a Complex World, 2000

Back to the catalog