Decoy Effect Pricing: The Bias That Barely Replicates (And Still Costs You Money)

Across 91 separate attempts covering 23 product categories, marketing researchers Sybil Yang and Michael Lynn could only reproduce the decoy effect 11 times. That is a 12% hit rate for one of the most confidently repeated ideas in popular behavioral economics.

If you have read a business book in the last fifteen years, you know the story: companies plant a deliberately bad middle option so the expensive one looks reasonable, and you fall for it every time. The research on decoy effect pricing is a lot messier than that — the lab findings are real, the real-world findings mostly are not, and the practical lesson is the opposite of what most articles tell you. This post covers what the original studies actually found, what happened when researchers tested the effect on realistic products, and what to do at the pricing page instead.

The standard story about decoy effect pricing

The effect has a real academic pedigree. In 1982, Joel Huber, John Payne, and Christopher Puto published Adding Asymmetrically Dominated Alternatives in the Journal of Consumer Research. They ran choice experiments across beer, cars, restaurants, lottery tickets, films, and televisions, and found something that standard choice theory said was impossible.

The principle they broke is called regularity: adding a new option to a menu should never make an existing option more popular. It can steal share, or do nothing. It should not grow the pie for someone else. Huber and colleagues showed it could — by adding a third option that was clearly worse than option A on every dimension, but only worse than option B on some. That third option is the decoy, and it exists to make A look good by comparison.

The example everyone quotes came later. In Predictably Irrational, Dan Ariely described a set of subscription options modeled on The Economist and put them to about 100 graduate students:

Nobody picked the print-only option. Its entire job was to make print-plus-web look like a free upgrade — and preference flipped from 68/32 in favor of the cheap plan to 16/84 in favor of the expensive one. It is a clean, memorable demonstration, and it is the reason “watch out for the middle option” became internet-standard advice.

What happened when researchers tried it on real products

Here is the part the business books skipped. In 2014, the Journal of Marketing Research ran what amounted to a public reckoning.

Sybil Yang and Michael Lynn’s paper, More Evidence Challenging the Robustness and Usefulness of the Attraction Effect, reported 91 attempts to produce the effect across 23 product classes and 73 different decoyed choice sets. Eleven produced a reliable effect. Their conclusion was blunt: the conditions required are so restrictive that the practical validity of the finding should be questioned.

In the same issue, Shane Frederick, Leonard Lee, and Ernest Baskin published The Limits of Attraction, which found the effect held mainly when options were described by simple numeric indices, and faded when they were described with pictures, qualitative attributes, or realistic product detail.

Huber, Payne, and Puto responded in the same issue with Let’s Be Honest About the Attraction Effect, arguing the effect remains robust when the original conditions are actually replicated — the disagreement, in their reading, is about moderators rather than existence. That is a fair point, and it is also the whole problem for a consumer. Here is what the moderators look like in practice:

Synthesized from Yang & Lynn (2014) and Frederick, Lee & Baskin (2014), Journal of Marketing Research.

Why decoy effect pricing still costs you money

The replication trouble does not get you off the hook — it just relocates the risk. Three things are worth separating.

First, the domain where the effect does replicate is exactly the domain where you spend recurring money. Software subscriptions, cloud storage, streaming tiers, and phone plans are the purest possible version of the lab setup: two or three numeric attributes, presented side by side, with dominance visible in a second. That is where the effect is strong and repeatable. A restaurant menu is not that. Your SaaS pricing page is.

Second, the recurring structure multiplies a one-time slip. A decoy that talks you into a $9-per-month upgrade instead of a $5 one is a $4 mistake exactly once, and a $48-per-year mistake forever. Subscription pricing converts a single moment of comparison into a permanent line item, which is why running a periodic subscription audit that surfaces forgotten charges catches more money than trying to out-think each pricing page in the moment.

Third, the decoy is rarely the main event. The heavier lifting on most pricing pages is done by the reference number set before you compare anything — the crossed-out “was $199,” the “most popular” badge, the annual price displayed as a monthly figure. Those are anchoring and framing effects, and unlike the decoy, they replicate broadly. If you want the highest-yield defense, it is understanding how framing effects shape the price you see and how anchoring sets your reference point before you start, not hunting for decoys.

I spent a while genuinely enjoying the decoy explanation — it is a satisfying idea for anyone who likes systems, and as a software engineer I’ve built the exact kind of three-column pricing table it describes. Then I read the 2014 exchange and had to update, which was annoying in the useful way. The version I kept is narrower: when I’m looking at a pricing table with clean numeric tiers, I assume the layout is doing work on me. When I’m looking at anything with photos, adjectives, or ten different features, I assume the decoy story is decoration and the real pressure is somewhere else. That distinction has saved me more money than the original insight ever did.

A worked example: what one nudged tier costs over five years

Abstract percentages do not change behavior. Arithmetic does. Take a three-tier pricing page of the exact shape the research says works — numeric attributes, side-by-side layout, an obviously dominated option in the middle:

The middle tier is there to be rejected. It makes Pro look like double the storage for one extra dollar, which is a genuinely good deal relative to Plus — and an irrelevant comparison if 200 GB was always enough. The decision that actually matters is Basic versus Pro: $480 over five years, or $960 if you make the same call on two services.

Note what the arithmetic does that the bias framing does not. It moves the comparison from “which tier is best value” to “what am I giving up $480 for.” Those are different questions, and the second one is answerable. Anyone who has run the numbers on what a decade of small recurring charges compounds to already knows the shape of the answer.

When the standard advice is right

“Beware the middle option” is not useless. It is a good heuristic in a specific, identifiable situation — and recognizing that situation is most of the skill.

The advice holds when all three of these are true at once: the options differ on two or three quantitative dimensions, they are laid out for direct side-by-side comparison, and one option is obviously worse than another on every dimension. If a tier costs the same as a better tier, or costs more for strictly less, you are looking at a real decoy. Pick on absolute value and ignore it.

The advice is close to useless when you are choosing a car, a house, an insurance policy, a restaurant entrée, or anything with photographs and prose. Under those conditions the published attempts largely failed to find the effect. Spending your attention there is spending it in the wrong place — and attention is the scarce resource, which is the same reason friction-based systems beat willpower for impulse buying. Rules that fire automatically outperform vigilance you have to sustain.

Four questions to ask before you buy the middle tier

1. What would I have picked with only two options? Mentally delete the tier nobody would rationally choose and re-decide. If your answer changes, the layout moved you, not the value.

2. What is the annual cost, in dollars? Tier upgrades are quoted monthly because $4 sounds like nothing. Multiply by 12 before you compare. A three-tier page where the gaps are $5 and $9 a month is really a page where the gaps are $60 and $108 a year.

3. Which specific feature am I buying, and when did I last need it? Name it. If you cannot name a single feature in the upgrade you used in the past 90 days, the upgrade is aspirational — you are paying for the version of yourself you plan to become. That is closely related to the ownership premium that makes people overvalue what they already have, applied forward in time.

4. Can I start one tier down and upgrade later? Almost always yes, and almost nobody does it. Downgrading requires admitting a mistake; upgrading feels like a reward. Start low and let real usage make the argument. The asymmetry in how those two moves feel is the single most exploitable thing about you on a pricing page.

What to take from a bias that only half-works

There is a broader lesson buried in this one. A large amount of popular personal finance content is built on single studies from the 1980s and 1990s that were later contested, narrowed, or failed to replicate. The decoy effect is unusual only in that the challenge and the defense were published side by side, in the same issue, by the original authors and their critics — which is science working properly, in public.

The practical response is not cynicism about behavioral economics. It is calibration. Effects that require lab conditions belong in the specific settings that resemble lab conditions. Effects that show up everywhere — anchoring, framing, the friction of payment, the pull of defaults — deserve standing rules in your financial life. Sorting one from the other is worth more than memorizing a longer list of biases.

Photo by R.D. Smith on
Unsplash

Written by Chris Steve

Chris Steve is a software engineer with a deep interest in personal finance, behavioral economics, and AI. He started Money & Planet to share clear, research-backed money guides — the kind that explain the math instead of pushing products. His writing focuses on long-term wealth building, the psychology behind spending and investing decisions, and the practical tools regular people can use to make smarter financial choices.