Skip to content

Trace Decay Review

Monitor — instantiates Revealed-Use Path Alignment

After a redesign, checks two things at once — that the old desire path is actually fading, and that no worse path has popped up somewhere else.

A redesign that looks right on launch day can still have failed. Trace Decay Review is the after-the-change watch that asks whether the fix actually took, and it insists on two coupled signals rather than one. Did the old trace decay — did people genuinely stop cutting the former path? And did a new or displaced trace appear elsewhere — did the friction simply move? Its defining stance is that success is not "old path down" on its own; it is "old path down and nothing worse popped up." The counter-trace search is what separates a real fix from mere displacement. Where a before/after snapshot measures the direct effect at the target site, this review is longitudinal and hunts sideways for the side effect.

Example

A company's employees kept working around a buried expense-policy page by direct-messaging one particular finance staffer — an informal route that quietly made her a single point of failure. The team redesigned the help center so the policy sits front and center, and set a Trace Decay Review running monthly against support metrics. It watches two curves at once. First, the old trace: are direct messages to that staffer actually declining month over month, toward the pre-workaround baseline? They are — good. But the review's second, decisive job is the displacement hunt: did the volume just relocate? Scanning adjacent channels, it catches a small new spike of "where do I file this?" searches pointing at a second, neighboring gap. So the win is booked and the next friction point is queued — instead of the team declaring total victory off the one curve that happened to move the way they hoped.

How it works

What distinguishes this monitor from a single after-the-fact reading is that it is longitudinal and displacement-seeking:

  • Watch the old trace decay over time. Not one after-snapshot but a curve — does the former path keep falling toward baseline over weeks and months, or dip and rebound?
  • Actively search for displacement. Scan adjacent channels, queues, and emergent shortcuts for a new trace, rather than only re-measuring the original site.
  • Couple the two signals. A pass requires both: the old trace gone and no worse trace taking its place.

Tuning parameters

  • Decay window — how long to watch before calling the old trace "gone." Too short and a novelty dip gets mistaken for a durable fix.
  • Displacement search breadth — how many adjacent channels get scanned for a new trace. Scan narrowly and you miss the mole popping up one channel over.
  • Decay threshold — how close to baseline counts as success. A lax threshold books wins that later rebound.
  • Baseline reference — what "faded" is measured against. A wrong or flattering baseline makes a mediocre result look like a win.
  • Cadence — a one-off post-mortem or a standing monitor. Because workarounds can re-emerge, a standing watch is what catches a late rebound.

When it helps, and when it misleads

Its strength is that it keeps the whole archetype honest at the end: a redesign is not validated by good intentions or by a single favorable after-reading, but by the old trace actually dying and nothing worse taking its place. The displacement hunt is the specific guard against the most common self-deception in this work — declaring victory when the behavior has merely moved.[1]

Its failure modes track its dials. Too short a window mistakes a novelty dip for durable decay. Too narrow a displacement search certifies success while the workaround thrives one channel over. And, run backwards, it degenerates into a victory-lap dashboard that reports only the metric the redesign was always going to move. The discipline is to pre-commit the decay threshold and window, scan widely for displacement, and keep watching for a rebound rather than closing the file after one good reading.

How it implements the components

Trace Decay Review realizes the after-the-change side of the archetype's machinery — the monitoring parts:

  • post_alignment_monitor — it tracks whether the old trace fades and the new official route is genuinely used once the change lands.
  • counter_trace_search — it actively hunts for a displaced or newly-emergent trace elsewhere, so a "fix" that merely relocated the friction is caught rather than celebrated.

It does not perform the change it evaluates (Route Closure with Alternative, Informal Route Legalization Patch) and it does not run the direct before/after magnitude comparison at the target site — that paired snapshot is Before/After Trace Monitoring.

Notes

The displacement half is what separates this from a plain before/after comparison — and it is why a narrow-scope decay review is worse than none. Reporting only the old curve, without scanning for where the behavior went, manufactures false confidence: the metric drops, the problem persists somewhere unwatched, and the file gets closed.

References

[1] Problem displacement — the effect, well known in situational crime prevention, whereby blocking a behavior at one location or channel pushes it to another rather than eliminating it. It is the precise failure the counter-trace search exists to detect before a redesign is declared a success. ↩