← Back to blog
Local Marketing5 min readNovember 11, 2026

One Cent or Free? What the Chocolate Experiment Means for Restaurant Rewards

At an MIT booth, both chocolate prices dropped by one cent, and the share of people choosing a Hershey's Kiss went from 14% to 42% once it became free. Here's the study, its limits, and a careful way to test free rewards with your own guests.

Drop a price by one cent and you'd expect almost nothing to change. Drop it from one cent to zero, and something different can happen.

That's the question behind a well-known field experiment with two chocolates and a booth on a university campus. It's worth knowing if you're planning a "free" reward, a freebie promotion or a stamp card for your guests.

The experiment: two chocolates, one booth

In 2007, Kristina Shampanier, Nina Mazar and Dan Ariely published "Zero as a Special Price: The True Value of Free Products" in the journal Marketing Science. In their Experiment 2, they set up a booth in MIT's student center:

  • Two boxes of chocolates, a Hershey's Kiss (the lower-value option) and a Lindt truffle (the higher-value option), with a large sign: "one chocolate per person."
  • The price of each chocolate was on a small sign lying flat on the table, readable only up close.
  • These were real purchases, not a hypothetical survey.
  • People who looked at the prices but didn't stop or buy were recorded as choosing "nothing."

The two price pairs we focus on:

  • Cost condition: Kiss 1¢, Lindt 15¢.
  • Free condition: Kiss 0¢ (free), Lindt 14¢.

Both prices dropped by exactly one cent, so the 14¢ gap between the two chocolates stayed the same. A third condition (Kiss free, Lindt 10¢) was also run; in total, 398 people took part across the three conditions.

The result: the free option pulled ahead

Among all recorded choices, including people coded as choosing nothing:

  • Hershey's Kiss: 14% → 42%
  • Lindt: 36% → 19%
  • Nothing: 50% → 39%

A one-cent change on both items, and the share choosing the Kiss tripled once it was free. In a standard cost–benefit view, cutting both prices by the same cent shouldn't shift people from Lindt to the Kiss. The authors call this pattern the zero-price effect: a free item seems to be valued as more than just "one cent cheaper."

The authors also report no significant difference in demand for the Kiss between the two free conditions (Lindt at 14¢ or at 10¢).

A note on what these numbers are: they are shares of chocolate choices at a booth, not sales revenue, profit or repeat visits.

The limits worth knowing

The authors are open about the design's limits, and so should we be:

  • People were not individually randomized. The price signs were switched roughly every 45 minutes (30-minute sessions with 15-minute breaks). Different people saw different conditions; it's not the same person changing their mind.
  • "Nothing" is imperfect. Some passers-by coded as "nothing" may not even have noticed the offer, which the authors acknowledge.
  • It's chocolate, not restaurants. The study did not test restaurant guests, customer retention, loyalty programs or revenue. Nothing in it promises that a free reward will bring guests back.

What it does not tell a restaurant

"Free" clearly changed what people picked at that booth. It doesn't tell you whether a free dessert brings a guest back next month, or whether the reward pays for itself. A reward that gets claimed a lot can still lose money, and one that gets claimed less can still be worth it.

Our practical hypothesis: measure take-up and margin

This part is our application, not something the study tested. If "free" draws extra attention, a free reward may get claimed more than a small discount of similar cost. Whether that's good for your restaurant depends on what it costs you. Before you assume a freebie works, track two things:

  • Take-up: how many guests claim or redeem the reward.
  • Margin: what each redemption costs you, and whether those guests come back or spend more.

A few low-cost ways to compare options:

  • Free item vs. small discount. For example, a free coffee versus a modest discount on the next visit, with a similar cost to you.
  • Low-cost free items. If "free" is what draws attention, pick a reward that's cheap for you to give.
  • Clear rules. State any conditions or expiry dates plainly, so guests know exactly what they're getting.

And one firm rule: keep perks separate from reviews. Never offer a reward in exchange for a review, and never ask only happy guests to leave one. Rewards are for visits and purchases; reviews should be honest and unprompted by incentives.

How to test it with your own guests

Because the study can't predict what will happen in your restaurant, treat any change as a small experiment:

  1. Pick one change (for example, swapping a discount reward for a free item of similar cost).
  2. Write down what you'll watch before you start: redemptions, cost per redemption, repeat visits from members.
  3. Compare a period before and after, and keep everything else as steady as you can.
  4. Keep it or drop it based on your own numbers, not on the chocolate study.

Where Westify fits

Westify's loyalty card lives in Apple Wallet and Google Wallet, so guests can see their stamps on their phone instead of carrying a paper card. If you test a reward, it's a simple place to run it. See the loyalty card (westify.app/loyalty-card).

Westify also replies to your Google reviews and turns them into reports on what customers say, so recurring themes, including comments about rewards, are easier to spot. Replies are drafted in your voice. If you're starting with reviews, see how to respond to Google reviews (westify.app/blog/how-to-respond-google-reviews-complete-guide) or try our free tools (westify.app/tools).

Fair caveats

  • This was a field test with chocolates at a university booth, not a study of restaurants, loyalty programs or customer retention.
  • The percentages are shares of all coded choices (including "nothing") in Experiment 2; conditions alternated over time rather than being randomized per person.
  • The restaurant ideas above are our hypothesis to test. Results will vary by restaurant, guests and program.

Source

- Shampanier, K., Mazar, N., & Ariely, D. (2007). Zero as a Special Price: The True Value of Free Products. Marketing Science, 26(6), 742–757. https://doi.org/10.1287/mksc.1060.0254 (Experiment 2, Figure 3)

Start with Westify

Plans from $29 a month

Get started