The decoy price: why the famous trick is shakier than it sounds

The idea is old and often repeated. You offer two options, say a 2-pack and a 6-pack. You add a third that is clearly worse than the 6-pack: nearly the same price, fewer items. Shoppers, the story goes, now choose the 6-pack more often. Researchers call this the attraction effect, or the decoy effect, and it is a staple of pricing advice.

A paper in the Journal of Marketing Research, “The Limits of Attraction” by Shane Frederick, Leonard Lee and Ernest Baskin, tested how far the effect reaches. Here is what they report.

The famous example did not replicate. The authors quote a well-known classroom result in which adding a decoy raised the share choosing the target option from 32% to 84%. They write that they failed to replicate several of the most-cited results, including that one.

It showed up in number tables and mostly nowhere else. Across 38 studies, they found significant attraction effects in four of five cases where products were shown as a table of numbers, such as a durability score of 7.2 and a cleaning score of 5.5. In 27 studies where at least one attribute could be experienced directly, such as drinks people actually tasted, tissues they touched, or apartment views shown as photos, they found no significant attraction effect, and one significant effect in the opposite direction.

Their conclusion is blunt. They say the boundary conditions seem so restrictive that the effect’s practical validity should be questioned.

The other side

This is a debate, not a settled verdict. Joel Huber, John Payne and Christopher Puto, who first described the effect, replied in the same journal (Journal of Marketing Research). They argue that the effect stays robust whenever study conditions are closely replicated, and they discuss factors that weaken or strengthen it. We read only the abstract of their reply, so we cannot say how they answer each of the replication failures.

What this means for a store

Our reading, not the paper’s claim: a store is much closer to the photo-and-product case than to a table of numbers. Shoppers see pictures, read descriptions and know what a 6-pack is. That is the situation where the paper found the decoy did nothing.

• Do not add a deliberately bad option on the strength of “the research says so”. The research is split, and the part that held up looks least like a real product page.
• If you do try one, treat it as a normal A/B test with a clear metric, such as revenue per visitor, and be ready for it to do nothing. A decoy can also confuse shoppers or look manipulative. That is a risk we are naming, not a finding from these papers.
• A bigger bundle may sell for ordinary reasons: a lower price per unit or a clear photo of what is inside. Those are easier to test and defend.

What to do this week

1. Look at your current options. If one exists only to make another look good, ask whether a shopper would ever buy it.
2. Pick one product with several pack sizes. Write down, in one line each, why a customer would choose each size.
3. If the lines for two sizes read the same, merge them or change one so the difference is real.
4. Only then, if you want to test a decoy, note the starting numbers and decide how long to run it before you begin.