Nudge Wiki

Concept

Priming

Also known as: Semantic Priming, Spreading Activation

priming associative-memory advertising

Exposure to a cue (a word, image, sound) activates a related "node" of knowledge in memory; because memory is densely interconnected, that activation spreads and raises the salience of everything else associated with the cue — making related thoughts, judgments, and choices more likely immediately afterward.

Psychologist Adrian North uses this "spreading activation" framing to explain why playing French music near a wine display doesn't just create ambience — it activates "French" associations broadly, and consumers become more likely to reach specifically for French wine (see Supermarket Wine and Music Study). He extended the same logic to advertising: in a follow-up study, adverts whose background music was thematically linked to a specific product feature produced better recall of that feature than adverts where the music didn't match, turning the mechanism into "a direct tool that's theoretically driven" for ad design rather than just an interesting curiosity.

Two further examples are mentioned in passing without a measured outcome: British Airways pairing its brand with classical music, and Guinness linking itself to rugby and St. Patrick's Day.

The mechanism also runs in reverse: a visible currency symbol on a restaurant menu primes "spending" itself as a concept, making diners more price-conscious. Removing the dollar sign entirely raised spend by about 6% (see Yang Dollar Sign Menu Study).

Phil Barden extends priming to a person's active goal state, not just external cues: hungry participants fitted with eye-tracking glasses looked significantly more at food signage than non-hungry participants, despite neither group being aware of the difference (see Hungry Eye-Tracking Goal-Priming Study). Ben & Jerry's found ice-cream purchase intent rose specifically when it was raining, and rain-triggered digital ads doubled banner-click volume (see Ben and Jerry's Rain-Triggered Ad Campaign). Barden even cites a physiological mechanism: when thirst activates a learned Coca-Cola/refreshment association, receptors in the optic nerve for perceiving red become measurably more sensitized, making a red Coca-Cola can easier to spot (see Coca-Cola Thirst Red-Perception Study). Physical presentation primes valuation too — participants were willing to pay 60% more for a fast food item shown physically rather than only in a photo or menu description (see Fast Food Physical-Presentation Willingness-to-Pay Study).

A money-priming demonstration shows the mechanism working on purchase intent directly: students shown a sofa for sale against a background of pennies and dollar-bill shading declared purchase intent at 56%, versus 36% for students shown the identical sofa against a neutral cloud background (see Mandel and Johnson Money-Priming Sofa Study).

Priming operates on perception directly, not just judgment (Episode 196). Dutch researchers played participants pure white noise while telling them it contained a hidden segment of Bing Crosby's "White Christmas" — roughly a third of listeners reported actually hearing the song, which wasn't there at all (see Bing Crosby 'White Christmas' Expectation Study). Priming is also cited as a genuine, well-evidenced phenomenon in Episode 195, which contrasts it directly with Diederik Stapel's fabricated "metaphorical priming" research — real priming findings (like the studies above) are distinguished explicitly from Stapel's fake ones, and from three studies the host of this podcast deliberately invented as a live demonstration mid-episode (see Nudge Podcast Fabricated-Studies Demonstration — not real research).

A full episode devoted to testing priming's reliability, with five failed replications (Episode 245). Daniel Kahneman's Thinking, Fast and Slow cites priming across roughly 35 separate points, including a chapter titled "The Marvels of Priming." Much of that chapter's evidence has since fallen apart: researcher Doyen failed to replicate one of its most prominent studies (see Doyen Florida-Effect Replication Failure Study (2012)), and a follow-up analysis by the site Replicability Index found 11 of the 12 studies Kahneman's priming chapter cites to be unreliable. Kahneman himself responded in a 2012 open email to the priming research community, writing that while he remained a general believer in priming, he feared it was "a train wreck looming" — and it largely was.

Phill Agnew ran his own five-part, informally-controlled replication test with over 100 British participants recruited via VoxPopMe, attempting to reproduce five separate well-known priming findings. None replicated as originally reported: a politeness-word prime ("authentically") failed to raise self-rated authenticity and if anything moved participants in the opposite direction (see Bargh Chen Burrows Politeness-Priming Study (1996)); an Apple-logo creativity prime produced significantly more creative brick-uses from participants shown the IBM logo instead, a statistically significant reversal (see Fitzsimons Apple-Logo Creativity Priming Study (2008)); a guilt-word prime raised reported fast-food purchase intent only slightly, with no statistical significance (p = 0.573) (see Goldsmith Cho Dhar Guilt-Priming Candy Study (2012)); a salad/cheesecake semantic-anchoring prime moved calorie estimates in the opposite direction from the original study in both conditions tested (see Chernev Semantic-Anchoring Cheesecake-Salad Study (2011)); and a reading-speed version of the famous "Florida effect" elderly-word priming study found no statistically significant difference in reading speed between aging-word and youth-word word lists (see Bargh Chen Burrows Florida-Effect Walking-Speed Study (1996)). Agnew concludes that reliable, well-replicated principles (loss aversion, anchoring, scarcity, the IKEA effect) should be preferred over priming-based tactics, which — however popular among marketing gurus — appear too weak an effect to reliably move behavior at any meaningful scale. See Replicability Crisis.

A direct follow-up revises the theory: it's contextual relevance, not stimulus size, that determines whether a prime works (Episode 253). Philip Graves, responding directly to Episode 245's five failed replications, argues Agnew's tests were fair replications of genuinely weak studies, not evidence priming never works — he defends several other priming findings as reliable, and offers a unifying explanation for the difference. Graves's reliable examples: an fMRI study finding the brain's reward centers show more activity when participants believe they're drinking a more expensive (but actually identical) wine (see Plassmann Wine Price fMRI Study); a companion restaurant study where diners rated an identical cheap wine, and the meal alongside it, as tasting significantly better under a premium-looking label (85% higher for wine, 50% higher for food when framed as "old world" vs "new world") (see Wansink van Ittersum Wine-Label Restaurant Study (2007)); a 2017 Oxford study finding wine opened with a corkscrew rated 10% higher in quality than identical wine opened with a screw cap (see Spence Wang Cork vs Screw-Cap Wine Study); Kahneman and Tversky's own founding anchoring demonstrations — multiplication-sequence estimation and "how many African countries are in the UN" — which Graves calls fully reliable, unlike some "parlor game" priming studies (see Kahneman Tversky Founding Anchoring Studies); Solomon Asch's 1946 trait-order impression study and Richard Shotton's modern vodka-description replication, both re-cited here (see Asch and Shotton Trait-Order Impression Studies); a study finding artwork displayed near cutlery made the identical cutlery seem more luxurious (see University of Georgia Art-Cutlery Luxury-Perception Study (2008)); the "new shoe smell" Nike scent study, now with its full citation (Hirsch, 1991) — 84% of shoppers more likely to buy Nike trainers displayed in a room with a light floral scent (see Nike, Malibu, and Novotel Scent-Marketing Cases); and a real-world consulting case where reordering a smoothie's front-of-pack ingredient list (fruit first vs. vegetable first) doubled sales of an identical drink (see Innocent Smoothie Ingredient-Order Priming Case).

Graves's proposed rule: a prime works when it's contextually relevant to the actual experience being judged — a wine's price or origin genuinely predicts how much you'll enjoy it; a shoe's smell genuinely relates to how fresh/new it seems; a smoothie's lead ingredient genuinely predicts its taste. A prime fails when the link is weak or arbitrary — a website's background color has no real connection to purchasing, and an elderly-associated word has no real connection to walking speed.

Agnew explicitly revises his own conclusion from Episode 245 in response. He maintains that all five of his original tests still failed to replicate, and that the specific studies he targeted remain unreliable. But he concludes his original explanation — that priming fails because the stimulus is too small or subtle — was itself wrong: contextually relevant primes (a wine's price, a shoe's scent) are just as small and subtle as the primes that failed, yet still work. The determining factor is whether the prime is meaningfully, causally connected to the judgment being primed, not how large or forceful it is.

Discussed in

  • Episode 1 — 1- How music affects what we buy
  • Episode 6 — 6- The psychology behind effective pricing
  • Episode 29 — 29- Why first impressions matter
  • Episode 42 — 42- Common pricing mistakes most businesses make
  • Episode 195 — 195-the-lying-psychologist-who-fooled-the-world (distinguishes real priming from Stapel's fabricated research)
  • Episode 196 — 196-7-cognitive-biases-that-might-ruin-your-life (Bing Crosby expectation-hallucination study)
  • Episode 209 — 209-can-i-persuade-20-strangers-to-listen-to-nudge (reframed as "pre-suasion"; beer/mouthwash superimposition; environmental-trigger studies)
  • Episode 245 — 245-i-debunked-psychology-s-greatest-myth (five failed replication attempts; Doyen; Kahneman's open letter)
  • Episode 253 — 253-did-i-get-it-wrong-revisiting-priming (Philip Graves's rebuttal; contextual-relevance framework; wine, cutlery, scent, and ingredient-order studies)

Related