What the decoy effect is
The decoy effect is a shift in choice caused by adding an option that few people want. Huber, Payne and Puto described it in the early 1980s: when they added an option that was clearly worse than one existing choice, but not clearly worse than the other, more people chose the option it was worse than. Researchers call it an asymmetrically dominated decoy: only one rival beats it on every count.
The reason is comparison. People find it hard to weigh two options that differ in several ways at once, such as price against features. A decoy hands them an easy contrast: one package is plainly better than the decoy for nearly the same money, so that package starts to feel like the sensible choice. The decoy does its work by being on the page; nobody has to buy it.
How the decoy effect shows up on a three-package page
Picture a website care plan sold in two versions. Essentials is low in price and covers monthly updates. Complete costs more and adds content edits and a monthly report. Buyers watching their budget lean to Essentials. Now add Plus, priced close to Complete, with updates and the report but no content edits. Beside Plus, Complete looks like a clear deal: a little more money buys a lot more.
Notice what makes Plus a decoy. It loses to Complete on value, yet it is not plainly worse than Essentials, because it includes more. That one-sided comparison is the whole mechanism. If the third option were simply a cut-down version of everything, or a far more expensive one, it would be an extra choice rather than a decoy, and it would not create the same easy contrast.
Building a fair three-package layout
The layout matters as much as the prices. Buyers compare packages row by row, so put the same features in the same order in every column, in the same units, and show what is missing as clearly as what is included. On a phone the columns stack, and Tversky and Kahneman's anchoring research suggests the first price people see colours how they read the rest, so decide which package a mobile visitor meets first.
Highlight one package only if you genuinely recommend it for most buyers, and give the reason in a line. A ‘recommended’ label on a plan you push for its margin is a small deception that sours the page. Keep the decoy honest as well: it should be a real package with a real use, such as a client who wants reporting but writes their own content, and you should be glad to sell it.
- The same features, in the same order and units, in every column
- A clear mark for what is not included, not empty space
- Package names that say who each one suits
- One honest recommendation, with its reason
- Prices shown the same way in every column, monthly or yearly, never mixed
Where decoys fall short or backfire
The original work gave people described options in a research setting. Real buyers bring budgets, business partners and competing quotes, and those can swamp a nudge from page layout. Treat the decoy effect as a reason to test a third package, not as a promise that one will lift revenue. With modest traffic a test may take months, so also note which package enquiries mention.
Decoys backfire when buyers notice the trick. A package priced so absurdly that nobody could want it reads as a set-up, and that suspicion spreads to everything else on the page. They also backfire when they add confusion: if a third column brings in features that appear nowhere else, buyers can no longer compare at a glance, and some will leave to think it over.
Try this with your own packages
Start with the decision you want buyers to find easy. If your middle package would serve most clients well, check whether anything on the page makes it look like a fair deal next to its neighbours. If nothing does, a carefully built third option may help. If your page already lists five or six plans, removing some may do more than adding a decoy.
Measure the share of buyers choosing each package and the revenue per order, not just clicks on a button. A layout that moves people to a bigger package but raises cancellations is not a win. Use the conversion rate calculator to estimate the value of a small rise in completed orders; that figure tells you how much design time the test deserves.
- Write one sentence on who each package is for, and merge any two that share a sentence
- Check that the package you recommend wins a row-by-row comparison with its neighbour
- If you add a third option, make it one real clients would choose
- Run the new layout long enough to see orders, not just visits
Where package pricing is heading
As more buyers ask AI assistants to compare providers, a package page will often be read as a table by software rather than viewed as a design. Clear rows, consistent units and plain names carry over well into those summaries. Tricks that depend on where a column sits carry over badly, because an assistant may list the packages in any order it likes.
- 01Package tables published as clean, structured data
- 02Package suggestions that adapt to what a visitor has already viewed
- 03Less patience from buyers for options that exist only to steer a choice
Questions & answers
It is when every option is real and fairly priced. A third package that suits some clients and helps others compare is honest design. A package created only to be rejected, at a price nobody would pay, is manipulation. A simple check: would you be comfortable explaining the layout to a client who asked why it looks that way?
Three is common because it shows a range and gives the middle option a neighbour on each side. Fewer can work for a simple offer. Beyond four, row-by-row comparison gets hard, especially on a phone. If you need many variations, keep three standard packages and handle unusual cases with a custom quote.
Close to the package you want to stand out, and a step below it in value. In many three-package layouts that means an option priced near the target but offering less. The position on the page matters less than the one-sided comparison: the decoy should lose clearly to one package and not to the other.
Compare the share of orders or enquiries for each package before and after the change, over enough time to cover normal swings in demand. Watch revenue per client and cancellations too. If the target package gains share but complaints or refunds rise, the layout is pushing people rather than helping them choose.
