Basic
€29per month
- 1 user
- no API access
Pro
€89per month
- 3 users
- no API access
Enterprise
€99per month
- unlimited users
- API access
Evidence
The original finding. Huber, Payne and Puto (1982) demonstrated the effect with 153 students across six product classes; the decoy raised the target option’s share by 9.2 percentage points on average.
Two large test programmes narrow it sharply. Frederick, Lee and Baskin (2014) did not find it once in the 27 studies in which at least one attribute could be directly experienced — but they did find it where every attribute was given as a number. Yang and Lynn (2014) arrived at the same pattern independently.
| Presentation | Share of successful attempts |
|---|---|
| purely numeric | 24% |
| with a picture or description | 4% |
The original authors objected — more narrowly than it sounds. Huber et al. (2014) hold the boundary conditions to be real and write themselves that the effect is probably rare in today’s market.
The older meta-analysis narrows it too. Across 92 comparisons, Heath and Chatterjee (1995) found a marked increase only for higher-quality target brands (+19.3 percentage points) and nothing dependable otherwise.
In the market it works strongly but rarely. Wu and Cosguner (2020) found in sales data from an online diamond retailer that the dominating option works when customers notice it — which, by the authors’ estimate, they do in only 11 to 25 percent of cases, depending on the price segment.
Limits
The effect needs attributes that are easy to compare and a comparison that is noticed — on current evidence, apparently attributes given as numbers. As soon as a photograph or a descriptive label is added, it disappears in the large test series. It also does not help every option equally: in the meta-analysis only the share of higher-quality target brands rose, and in a price-conscious sample the direction reversed. The market data are observational data from a different product field, and the dominance structure there arose incidentally from supplier prices, not from a pricing strategy. The original authors now see it similarly: Huber, Payne and Puto (2014) write that their 1982 paper was a demonstration testing an axiom, not a proposal for marketing practice. In a commercial data set of 586 people with some 4,000 matching choice sets, Huber himself found no gain for the dominating option.