A price tag is not a slip of paper that just says "how much." Its font, its spot, its color, the price sitting next to it — all of it bends the same number into a different decision. The figure 39 makes one person buy and sends another walking. From a 1995 University of Chicago study to 2024 neuroeconomics work at NeurIPS, here is how well seven retail pricing tactics actually held up in the data.

1. The left-digit effect — what $19.99 vs. $20 really buys

Start with a number that breaks your intuition. In their 2003 Quantitative Marketing and Economics paper, Anderson and Simester teamed up with a U.S. apparel catalog and randomized 23 identical items across $34, $39, and $44. Cheaper should sell better, right? Wrong. Sales at $39 ran 8% higher than at $34. Higher price, more revenue, same product.

This is the left-digit effect. Thomas and Morwitz's 2005 Journal of Consumer Research paper looked inside it with fMRI. Brains seeing $3.00 lit up the left-parietal price-processing region harder than brains seeing $2.99 — over a one-cent gap. The point: the brain re-computes the price when the leftmost digit moves, and largely shrugs off the rightmost one.

Swap the category, though, and the effect flips. Schindler and Kibarian (1996) found charm pricing lifting everyday grocery sales 5–10%. At the luxury end it does the opposite. Macciaroni's 2018 luxury-watch analysis found that prices like $9,999 read as "cheap-feeling" and dragged sales 12% below $10,000. A price signal always works alongside the neighborhood it sits in.

2. Dual-label sales — was that "was" price ever real?

Compeau and Grewal's 1998 Journal of Public Policy & Marketing meta-analysis pooled 27 studies on reference-price advertising. The verdict: willingness-to-pay rose an average of 16% when a "was" price sat next to the current one. Here's the unsettling part. Inflate that "was" anchor absurdly above the real market price, and the effect didn't shrink. It worked even when it should have looked like a lie.

That finding is exactly why U.S. regulators leaned hard on "fictitious pricing" in the 1990s. California's 1995 BPC §17501 bars showing a "was" price unless the item actually sold at it within the past 90 days. Korea's Fair Trade Commission drew a similar line in its 2016 update. Enforcement still lags reality — Top Class Actions logged 73 dual-pricing suits in 2024, clustered in apparel and home goods.

So the only thing to check is one question: when was this "was" price last an actual transaction? If something has been "on sale" all year, that "was" figure is a marketing stake in the ground, not transaction data.

3. The decoy effect — the Economist subscription table Ariely dug up

Say "decoy effect" and everyone reaches for this one. Dan Ariely's 2008 telling in Predictably Irrational, lifted from The Economist's subscription page:

• Online only — $59
• Print only — $125
• Online + print — $125

The middle one is a chump's choice — same money, but the third option throws in online for free. Yet with that useless option sitting there, 84% of students grabbed the combo. Ariely pulled the middle option and showed only two; combo dropped to 32%. One dominated decoy lifted revenue by roughly 43%.

Huber, Payne, and Puto's 1982 Journal of Consumer Research paper first named it: "asymmetrically dominated alternatives." Over the next four decades, 200+ replications held up, with target-option share rising 9–18 percentage points on average. The 2020 Sivakumar and Krishnan meta-analysis pinned two caveats: a decoy that's too obviously inferior loses its grip, and the effect bites hardest in categories that resist comparison — wine, fragrance, software tiers.

4. Bundling — where the savings actually live

A bundle lifts revenue two ways. It pushes more items per shelf-minute, fattening per-transaction margin. And it makes per-item comparison harder, easing the pressure to cut margin. Stremersch and Tellis's 2002 Journal of Marketing paper sized both separately and found that the "comparison friction" effect added roughly 1.6× more to bundle revenue than the direct discount did. More confusion than savings, in other words.

Which is why bundles aren't reliably cheaper. Consumer Reports cracked open 200 U.S. cable + internet bundles in 2019 and found 41% of advertised "bundle savings" cost more than buying the pieces separately — hidden equipment fees, padded channel packs, and contract terms buried in the price. The same showed up in airfare-hotel packages (Sabre Research 2022) and Korean mobile plans (KISDI 2023).

The defense takes a little work but it's simple: search each component on its own, add them up, and hold the sum against the bundle headline. Those 30 seconds come back as 8–15% in real savings — or a trap dodged.

5. Anchor products — why a $50,000 bag sits by the door

Tversky and Kahneman's 1974 Science paper first caught anchoring in the act. They spun a roulette wheel for a random number, showed it to people, then asked something totally unrelated — "what percentage of UN members are African countries?" The random number tugged the answers along with it. The group that saw 10 averaged 25%; the group that saw 65 averaged 45%. What does a roulette wheel have to do with the UN? Nothing. People drifted anyway.

That $50,000 bag at the front of the luxury floor isn't there to sell. It's a stake. The $8,000 bag two displays in suddenly sounds reasonable beside it. Northcraft and Neale's 1987 study did this to real-estate pros: same house, same task, but one group saw a $149,900 list price and the other $119,900. Their average appraisals diverged by 11%. Trained experts don't walk out of the anchor's reach either.

Online it slips in through sort order. Look at "price: high to low" once on Amazon, flip to "low to high," and the high figure stays lodged — a mid-tier item now reads as reasonable. Unless you deliberately reset the range, the anchor keeps firing.

6. Time framing — why "$29/month" feels lighter

Soman's 2003 Journal of Consumer Research paper measured time framing head-on. The same $400 annual gym fee was shown as "$400 per year," "$33.33 per month," or "$1.10 per day." Sign-ups ran ~40% higher under the daily framing than the annual one.

Two reasons. A small unit instantly compares to a familiar everyday expense — a coffee — so the burden shrinks. And the smaller the unit, the more annoying it is to total up the full-year cost in your head. People dodge effortful arithmetic by reflex, especially at the register.

That's why Netflix, telecoms, and gyms lean on it almost universally. The defense is the mirror image: see "$X/month," multiply by 12 on the spot. One multiplication snaps the decision back to its real weight.

7. Price precision — why an exact figure lodges harder than a round one

Janiszewski and Uy's 2008 Psychological Science paper is a bit of a surprise. When a listing was posted at $4,985,000 instead of $5,000,000, buyer counter-offers came in 24% lower on average against the round number. Put plainly: the precise figure was the better anchor for the seller.

Why? A round $5,000,000 reads as "a rough starting point." A jagged $4,985,000 reads as "a price someone actually worked out." So buyers answer the second with smaller adjustments. Mason et al. (2013) got the same pattern in cars, used goods, and eBay auctions.

Notice it in daily life and you stop letting "huh, that's oddly specific" do your thinking — and put the real variables first: market average, comparable listings, your own budget.

Does knowing help? — Wilson et al. 1996 answers straight

Here's the question everyone lands on: if you know the trick, are you safe? Wilson et al.'s 1996 Journal of Personality and Social Psychology experiment answers it flat out. Explicitly warning subjects about anchoring shaved the effect by only about 27%. More than 70% of the pull survived the warning.

Two things to take. One, drop the swagger — "I know about this" is not immunity. Two, the partial cut is real. Knowing is worth its 0.27.

A six-step checklist for everyday shopping

1. Read the leftmost digit first. $19.99 lives in the $19 band, not the $20 one. Focus there, ditch the tail.
2. Doubt the "was" price. Ask whether it actually sold at that price in the last 90 days. Apparel, electronics, and furniture inflate it most.
3. Doubt the middle option. Three tiers where the middle looks "obviously bad" is the classic decoy setup. Decide between the outer two only.
4. Unbundle. Split the package into pieces and sum them. Thirty seconds, 8–15% expected savings or a trap avoided.
5. Convert the time unit immediately. "$X/month" → ×12. "$X/day" → ×365.
6. Sit on it 24 hours. For non-urgent buys, leave the cart and walk away. How that price feels outside the store is the most honest signal you'll get.