Nudge Wiki

Concept

Generation Effect (Audience Participation)

Also known as: Generation Effect, Fill-in-the-Blank Effect

memory advertising copywriting effort

Information a person has to partly generate themselves is remembered better than the same information handed over complete. The founding demonstration is deliberately small: show one group whole words and another group the same words with a letter blanked out (F_ST, D_G), and the blanked-out group recalls about 15% more of them (see Graf Slamecka Generation Effect Study). Richard Shotton re-ran the study on brand names with 415 people and got the same direction — 92% recall for the generating group versus 81% for the full-word group, a 14% improvement (see Shotton Trahan Brand Recall Generation-Effect Study).

Cancer Research UK built a 2019 outdoor campaign directly on this, running "OB_S__Y is a cause of cancer" on bus stops and billboards rather than spelling the word out (see Cancer Research UK Blanked-Letters Obesity Ad). A 2020 Leo Burnett/Shotton study applied the same blanked-letters device to brand names in ads (e.g. "HS_C" for HSBC), finding 14% higher recall than the spelled-out version (see Leo Burnett Shotton HSBC-Blanks Study (2020)). Specsavers took the same completion effort in a different direction, deliberately jumbling its own slogan on a billboard ("you should have gone Savers to Spec") to force active unscrambling (see Specsavers Jumbled-Slogan Case).

The counterintuitive part is that it cuts against marketing's default instinct. Shotton's framing: marketers assume a message must be made as simple and effortless as possible or the audience won't process it — but a message that costs the audience nothing to process ends up not being very memorable either. This is a genuine tension with the "make it easy" school of nudging, not a contradiction of it; see EAST Framework and the ease-versus-outcome problem discussed in Choice Architecture (Salience & Placement). Shotton resolves it as two different jobs: to change behavior, remove friction; to change perception of your product, consider adding a little.

Removing letters is only the crudest application. The underlying principle is giving the audience a small puzzle to solve. Shotton's favourite example is David Abbott's Economist ad — "I never read The Economist. Management trainee, aged 42" — which never states its (fairly nasty) claim that non-readers fail, but makes the reader assemble it, which both makes it stickier and makes the insult palatable (see Economist Ad Generation-Effect Puzzle Case). He argues a study should be treated as "a springboard, not a straitjacket": take the insight, then apply creative thinking rather than forcing the literal experimental setup onto your ads.

Asking a question instead of making a statement is the same mechanism in a different form. An ad phrased "Did you know wearing Avanti shoes can reduce your risk of arthritis?" was rated 14% more positively than the identical claim as a flat statement (see Ahluwalia Burnkrant Question-Framing Ad Study). Phill Agnew replicated this on paid Reddit ads for this podcast: changing "Ditch boring business podcasts" to "Bored of boring business podcasts?" made the ad 15% more effective (see Nudge Podcast Question-Framing Reddit Ad Test). Agnew also tested obscured words on his own episode promotion, getting 13% more clicks (see Nudge Podcast Generation-Effect Promotion Tests).

The related Information Gap (Curiosity Gap) explains why an unanswered question also drives action, not just recall — Agnew cites the Zeigarnik effect (we're more likely to act on something left unresolved) as his reason for expecting the obscured-name episode title to lift listens.

Shotton links the effect directly to the IKEA Effect: both are about effort raising value, but they solve different problems — the generation effect is about memory, the IKEA effect about appreciation.

Re-cited in Episode 234's memory-focused treatment, distinguished from the closely related Testing Effect (Error-Driven Learning): the generation effect is about generating information yourself during encoding (filling in a blank raises memory of what fills it); the testing effect is about retrieval itself — even a wrong guess — strengthening a memory already formed. Charan Ranganath re-cites the HSBC blanked-letters study's exact figures (see Leo Burnett Shotton HSBC-Blanks Study (2020)).

Two further real-world examples via Adam Ferrier (Episode 272). A UK blood-donation service ran an ad with every letter "O" removed ("We're running -ut -f O-type bl--d"), forcing readers to fill the gap themselves (see UK Blood Service O-Type Blanked-Letters Campaign). Specsavers has also run a series of deliberately jumbled airport-signage ads (e.g. "Welcome to Melbourne" outside a Sydney airport, and vice versa), inviting audiences to spot the mistake.

Discussed in

  • Episode 136 — 136-today-s-guest-is-ri-h-rd-sh-tto
  • Episode 162 — 162-the-best-of-nudge-in-2023 (year-end recap)
  • Episode 188 — 188-how-adverts-use-psychology-to-win-you-over (re-cites question-framing and blanked-letters studies)
  • Episode 234 — 234-one-simple-trick-to-improve-your-memory (distinguished from and connected to the testing effect; HSBC study re-cited)
  • Episode 272 — 272-the-nudge-that-persuaded-aussies-to-stop-speeding (UK Blood Service blanked-letters campaign; Specsavers airport-signage ads)

Related