Too fine a grain drowns turning points; too coarse hides breaks

What it is

One map has two hundred clicks. Another says only Learn—Buy—Use. The first drowns the step that actually changed direction among look-alike actions. The second swallows a break such as “OTP expired, start the form again” into a giant stage. Journey map granularity is set so turning points stay visible: where a person changes goal, changes channel, enters an irreversible stretch, or cannot recover unaided after failure. Too fine and too coarse are the same ruler pointed at the wrong object, not a taste issue.

Why it happens

Fine grain comes from treating interface steps as journey steps—every field, every confirm occupies a cell, visual weight flattens, decision-makers see only length. Coarse grain comes from funnel language that existed before the observation; exceptions that will not fit three bins are squeezed out. A turning point is defined by asymmetric consequence: go wrong and you cannot return, must change person, must change day. Grain should follow consequence, not screen count or department count. One map may still have two layers: stages for orientation, expansion only near turns, rather than one scale for the whole sheet.

Studying it

Ask people who did not draw the map to mark “if you could change only one place, which?” If answers scatter across many similar micro-steps, the grain is too fine. Then pin real failure cases onto the map and see whether the break lands on an existing step; if the answer is only “somewhere in Buy,” the grain is too coarse. Compare a fixed three-stage template with a consequence-cut map on hit rate for locating a fixable break. Outcomes: share of turning points that exist as their own steps, and the rate at which failure cases can be pinned.

Where it stops holding

Training material or a script may need a finer action layer than a decision map; that is a different artifact and should not be written back into the decision map. Regulatory audits sometimes force every disclosure click onto the page; fine grain is then a compliance demand and should be labeled so it is not read as experience priority. In highly practiced repetitive work, the micro-step itself is the failure (wrong account selected); fine grain is then reasonable. Forcing one set of stage names across cultures or product lines crams distinct turning points into a shared Use.

Applying it

• List positions with asymmetric consequence (irreversible, change of person, change of channel, no self-repair); those must be steps of their own.
• Merge consecutive look-alike micro-actions until merging would eat one of those positions.
• Stage names are an upper index only; lower steps use verbs from observation, not funnel words.
• Trial-pin two or three known failures: if they will not stick, add a step; if they stick but sit among a long run of indistinguishable cells, merge.

Related

• Same group: Q4.04.1 A journey map spans touchpoints and shows the whole course · Q4.04.2 Emotion and pain on the map need a source · Q4.04.3 An unverified map is a team hypothesis · Q4.04.4 As-is maps what happens; to-be maps what is wanted — keep them apart · Q4.04.6 Maps that cross departments need shared upkeep · Q4.04.7 An unrefreshed map drifts from a changed process
• Adjacent: Q4.07 Task analysis · Q4.05 Service blueprints
• Search terms: journey map granularity · turning point · experience breakpoint