A globally emphasised UI makes users immune to all emphasis

What it is

A red dot once brought people back to messages. After the home lights nine dots, three unread strips and a “must see” chip at once, the dot is decoration. People say “those are always on.” When “emphasised” is the base rate, emphasis no longer predicts a new consequence, and reaction drops to the whole class of cues, not to one badge. Immunity is what gets learned after accumulation, bound to time, repetition and undifferentiated outcomes. One author bolding a screen once produces no first place in that moment; there has not been time to grow immunity.

Why it happens

Attention treats a class of stimulus as a signal when it is relatively scarce and correlated with something that then must be handled. Global emphasis lifts the base rate of “red, motion, badge” toward every open, and the correlation thins: opening often yields nothing new, or something smaller than the appearance. The class becomes background texture. Habituation speeds this, but immunity is wider than repetition of one item: even a never-seen new device (another kind of motion) is filed under “the product shouting again,” because the category is already “the product is shouting,” not “this pixel changed.” Stacking a new device on an already immune surface should be predicted to fail, not to “catch them this time.”

Studying it

Follow one user’s click or mention rate for the same class of cue across week one, week four, week twelve, holding appearance constant. A drop with unchanged appearance is a dropping response. After immunity has formed, introduce a never-used emphasis device and see whether its first-time efficiency is already below the same device on a clean product — a test of category generalisation. Independent variables are emphasis base rate and whether a real consequence follows; the dependent is the response curve. Short lab repetition only shows the start of habituation; field weeks show immunity.

Where it stops holding

Rare signals that always carry a consequence you cannot ignore (payment failed, a security sign-in) can still work on an immune surface, because consequence correlation remains. Using those signals for daily campaigns spends the last exemption. New users in week one are often not yet immune; proving “the dot works” on their clicks overestimates retained users. Switching the cues off for a while can partly restore response; treating that as a licence to rinse users on a cycle will spend the recovery window too.

Applying it

• On an already immune page, stop experimenting with new devices. Cut base rate first: dots only for untreated items with consequences; campaign chips get a removal date.
• Write “no appearance without a consequence” for each class of cue. A consequence-free dot trains people to ignore every dot.
• Protect incident and security channels. Campaigns may not reuse their appearance.
• Check: sample retained users, count home objects that “look like emphasis,” and see whether clicks on them over four weeks sit near zero. High count plus near-zero clicks is immunity. Run two weeks of base-rate cuts (consequence-bearing signals only) and compare clicks in weeks three and four. If they are still zero, the cut may have missed the category they already filed as “shouting” (motion still on), rather than shipping another chip.

Related

• Same group: F3.11.1 Teams each grabbing emphasis for their feature accumulate into global emphasis · F3.11.3 Repairing global emphasis means re-ranking, not stacking new cues
• Nearby: A5.16 Habituation and cue failure · F5.11.2 Hierarchy fails when accent colour is everywhere · E6.11 Badges and dots
• Search terms: habituation · learned irrelevance · badge blindness · emphasis immunity