The same funny loading message lands once, then curdles into irritation

What it is

A joke is the canonical one-shot stimulus: its effect is largely spent the first time it lands; later exposures leave a residue, and the residue is not neutral — it is irritation. The same quip or the same animated gag loses goodwill with every repetition, fastest on the high-frequency touchpoints users pass daily. Advertising research calls this wear-out: persuasion drops steeply with repeated exposure. Humor is more fragile than decoration, because what it burns is the punchline itself.

Why it happens

The engine of a punchline is unpredictability: a setup builds an expectation, the punchline violates it, the gap produces the laugh. The engine burns its own fuel — the moment you get it, the ending is permanently in memory; the second pass leaves no gap to resolve, only decoding cost: rereading a setup whose end you know, waiting for a laugh that will not come. That is also why dead humor lands below neutral rather than at neutral. Decorative pleasure (a smooth transition) fades into the background and its attention cost disappears with it; a dead joke still interrupts you at its trigger, still makes you process a punchline you already know, and still sends you away empty-handed. Every repetition is an invitation that can no longer be accepted, and being invited to laugh again and again reads as being condescended to — "this product is fun" flips into "this product thinks this is funny." High-frequency touchpoints amplify everything: a witty loading line or an intro gag replayed on every launch accumulates double-digit per-user exposures within weeks, while the punchline was spent on the first one.

Studying it

• Paradigms: within-subject repeated exposure is the standard design — the same jokes presented several times with a funniness rating each time; the decay is already steep by the second or third exposure. Advertising research contributes frequency–response curves from real campaigns (click-through and liking plotted against frequency), with higher ecological validity but no way to isolate humor from the rest of the ad.
• Variables: exposure count and exposure spacing are the usual independent variables; funniness ratings, annoyance ratings, and behavioral signals such as skip and close are the usual outcomes.
• Methodological cautions: lab studies compress repetitions into one short session, while interface exposure spreads over weeks and across changing contexts — the decay curves do not transfer directly. More decisive is the outcome measure: a funniness scale that bottoms out shows only "not funny," but the residual state after failure is "annoying," and the two are fused at the bottom of the scale. Capturing the negative residual requires annoyance ratings or skip/close behavior; funniness alone cannot reach it.

Where it stops holding

What decays is the specific joke, not a light register. A fixed quip replayed verbatim dies fastest; fixed form with rotating content (a pool of different lines) spreads each joke's exposure thin and lives an order of magnitude longer. The axis is per-user exposure count, not total exposure: an easter egg seen once by a million users is still fresh; a line seen twenty times by a thousand users is rotten. Low-frequency placement protects by the same logic — annual recaps, first-run onboarding, gags that must be deliberately discovered — frequency there naturally sits below the decay line. Whether a light tone sustained as brand personality wears the same way is a question about personality expression, distinct from the verbatim repetition of specific jokes.

Applying it

• Inventory every humor element that replays verbatim — fixed quips, the same animated gag, an intro joke on every launch — and annotate per-user exposure frequency; anything on a weekly mandatory path gets de-humored.
• Place humor only in low-frequency slots: first-run onboarding, annual moments, easter eggs that must be found — keep per-user exposure of each specific joke in single digits.
• Where a high-frequency slot genuinely needs humor, rotate a pool and give it a retirement mechanism; rotation only delays decay, and a pool that is too small means placing nothing at all.
• To validate: compare first-exposure and recent-exposure behavior for the same joke (skip rate, close rate, negative-feedback keywords); rising negative signals with exposure count mark it dead. A cheaper default: any fixed humor element on a weekly mandatory path is treated as already dead.

Related

• Same group: P1.06.1 Jokes do not belong in error and loss scenarios · P1.06.3 Humor is tightly bound to culture and context
• Nearby: P1.02.2 High-frequency delighters become interference · P1.09.2 Surprise is the source of delight and zeroes out once expected · A5.16 Habituation and message failure
• Search terms: advertising wear-out · hedonic adaptation · habituation · incongruity