Desire Path Walkthrough¶

Method — instantiates Revealed-Use Path Alignment

Traverse the designed route and the worn-in one side by side, on site, to make the official path explicit and mark exactly where lived use departs from it.

Desire Path Walkthrough is a structured first-person inspection: you physically walk (or click through) the route the design intended, then look for where feet, hands, or attention have worn a different route, and record the two together. The move that makes it this mechanism and not its siblings is that it holds the official route in one hand and the observed divergence in the other — establishing the baseline and the gap in a single on-site pass. It does not aggregate logs the way a heatmap does, does not ask users why the way a go-along does, and does not price the friction the way a mapping session does. Its whole deliverable is an annotated "here is where the design expected people to go, and here is where they actually went."

Example¶

A museum's curators designed a numbered gallery route, 1 through 12, meant to unfold a story. On a busy Saturday an evaluator does the walkthrough: first she walks the intended sequence and writes it down, then she watches real visitors. Most enter, skip straight to gallery 7 (the famous painting), double back, and cut through a service corridor to reach the café — a door propped open by a wedge that clearly lives there. She photographs the worn track in the carpet, marks the propped door on the floor plan, and flags the three points where observed flow leaves the numbered path. The output isn't a fix or a reason — it's an annotated map: designed route in one colour, observed route in another, three divergence hotspots circled, handed to the interpretation step.

How it works¶

Baseline first. Make the intended route explicit and written down before looking for deviations — you cannot spot a departure without the line it departs from, and the official route is often unrecorded until now.
Stand where the user stands. Traverse from the user's position and knowledge, not the designer's, so you meet the same dead-ends and missing signs they do.
Capture the gap as evidence. Photos, marked plans, counts at divergence points — concrete artifacts, not "it felt like people cut across."
Stop at the gap. It surfaces where lived use diverges and deliberately hands off why (to a go-along) and whether to allow it (to the filter).

Tuning parameters¶

Traversal fidelity — one quick reconnaissance versus repeated walks at different times and loads. More passes catch intermittent and peak-only paths but cost hours.
Observer stance — pose as a naïve first-time user or as an expert who knows the intended route. The naïve stance surfaces confusion; the expert stance surfaces where even the informed break the rules.
Substrate — physical space, a workflow, or a UI click-through; the discipline (baseline, then divergence) travels even as the terrain changes.
Recording grain — narrative notes versus marked plans versus counts. Finer grain feeds a quantitative comparison downstream but slows the walk.

When it helps, and when it misleads¶

Its strength is that it is cheap, fast, and forces the official route to be written down at all — and it catches embodied, spatial deviations that no log can see (the propped door, the cut corner, the ignored sign). It is the design cousin of a gemba walk: go to the actual place and see for yourself rather than reason from a diagram.^[1]

Its central weakness is that a single walk is a snapshot — you see the paths present the day you looked and miss intermittent, seasonal, or after-hours ones, and you see only the people who showed up, not those the design already turned away at the door. The classic misuse is walking to confirm a redesign already chosen, noticing only the deviations that flatter it. The discipline that guards against this is to walk before you have a favourite, at more than one time, and to send the "why" to a go-along rather than narrating it for the users yourself.

How it implements the components¶

intended_path_model — the walk's first act makes the designed route explicit and documented, giving every later trace something to be measured against.
intended_actual_path_comparison — walking both in one pass yields the annotated designed-versus-observed gap, with the divergence points located.

It does NOT produce a reusable, at-scale raw-trace record — that's Use-Trace Heatmap and Clickstream Deviation Scan — nor recover why the divergence exists (that's Frontline Go-Along Interview) or judge its safety and legitimacy (Safety and Accessibility Review).

Instantiates: Revealed-Use Path Alignment — the walkthrough supplies the baseline map and the first-hand gap the rest of the loop reads.
Sibling mechanisms: Frontline Go-Along Interview · Clickstream Deviation Scan · Friction Mapping Session · Use-Trace Heatmap · Workaround Inventory · Before/After Trace Monitoring · Trace Decay Review · Safety and Accessibility Review · Route Closure with Alternative · Temporary Paving Pilot · Informal Route Legalization Patch

Notes¶

The walkthrough observes; it does not interview. That boundary matters: the moment you start asking users to explain themselves, you are running a go-along, and the moment you validate broad fit rather than a specific observed deviation, you have drifted into general user research (the wider User Context Validation archetype) rather than this trace-first one.

References¶

[1] A gemba walk is the Lean practice of going to the actual place where work happens to observe it directly, rather than managing from reports — "gemba" being the real place. The walkthrough borrows its core discipline: see the trace in situ, do not infer it from a diagram. ↩