The Option You Never Meant to Sell
You walk into a café. Black coffee is $3, a caramel latte is $5. You take the latte without much thought.
Now imagine a third line on the board: a small drip coffee for $2. Suddenly the $3 feels like the sensible middle, and quite a lot of people who would have taken the latte take the black coffee instead.
Nothing about the $3 coffee changed. What changed is what it was sitting next to.
We compare; we don't evaluate
People are poor at judging value in isolation and good at judging it relative to something else. Psychologists call it the contrast effect, and it has two well-studied flavours that get mushed together in business writing.
Extremeness aversion, or the compromise effect. Given three options, people disproportionately pick the middle one. Not the cheapest — that feels like settling. Not the dearest — that feels indulgent. The middle feels defensible, which matters, because most purchases are decisions we expect to have to justify to somebody.
The decoy effect, or asymmetric dominance. Add an option that is clearly worse than one of the existing options but not clearly worse than the others, and you raise the share of the option that dominates it. The decoy isn't there to be bought. It's there to make its neighbour look obviously correct.
The most famous demonstration is a subscription page: web-only for $59, print-only for $125, print-and-web for $125. Almost nobody takes print-only — it is strictly worse than the bundle at the same price. But its presence makes the $125 bundle read as a bargain rather than as the expensive option, and the split shifts sharply towards it. Remove the pointless middle line and the same page sells a very different mix.
Watch it happen
Add an option nobody buys, and watch the others move
A shelf of televisions. Nothing about the first three changes when you add the fourth — but which one looks reasonable does.
Share taking the $1,200 50"
21%
the top of the range
Share taking the $700 42"
59%
the compromise
Average price paid
$725
Share taking the $1,500
—
not on the shelf
- Share of buyers
How this is modelled. Each television has a fixed standalone appeal, and anything that is neither the cheapest nor the dearest on the shelf gets a constant bonus for looking like the sensible middle. Shares are a softmax over the result. It reproduces the direction of the documented compromise effect so you can see the mechanism — it is not sales data and it is not calibrated to any.
Show the numbers
| Television | Price | Share without the $1,500 | Share with it |
|---|---|---|---|
| 32" basic | $300 | 20% | 20% |
| 42" smart | $700 | 59% | 59% |
| 50" OLED | $1,200 | 21% | 21% |
| 55" OLED | $1,500 | — | not on the shelf |
Three televisions, then four. The first three don't change — not the specs, not the prices, not a word of the copy. Only the shelf around them changes, and where the buyers land moves with it.
Where this shows up when you're the one selling
Pricing tiers. If you have three plans and you genuinely want people on the middle one, the ends of the range are doing most of the work. A top tier priced for a customer you rarely win still earns its place by defining what "a lot" means.
Anchoring in a negotiation. The first number in the room reframes every number after it. This is why experienced negotiators care so much about who opens, and why a wildly unreasonable opener is a real tactic and not just bad manners.
Positioning against competitors. "We save you ten hours a week" lands harder next to "they save you two" than it does on its own. The comparison isn't decoration — it's the thing that makes the number mean something.
Service moments. A hotel that upgrades you unprompted has done something small in absolute terms and enormous relative to the flat expectation you walked in with. Contrast is where a fixed budget buys a disproportionate impression.
The honest version of this advice
A lot of writing on this topic ends with a wink — use it before it's used on you. That's worth being straighter about.
There is a real difference between framing and manipulation, and it isn't subtle. Framing helps someone see a trade-off they'd endorse on reflection: here is what more money buys you, here is what it doesn't, decide. Manipulation engineers a choice the person would reject if they saw the structure — a decoy built to confuse rather than clarify, a "was" price that was never charged, a middle tier crippled for no reason but to push people up.
The practical test is simple: would you be comfortable explaining your menu design to the customer? If showing someone the mechanism would make them feel well served, you're framing. If it would make them feel worked on, you're not, and the fact that it converts is beside the point. Contrast effects are strong enough that the short-term numbers will look fine either way. The difference shows up later, in whether people come back.
Three things to actually do
- Look at your menu as a shape, not a list. What is the cheapest thing, what is the dearest, and what does that range say about the one in the middle? You are already sending this signal whether or not you meant to.
- Make the differences legible, not just the prices. A contrast only helps if the reader can see what the extra money buys. Three prices with vague feature lists produce confusion, and confused buyers default to the cheapest or to nothing.
- Test it, and mean it. Different orderings and tier structures are genuinely testable. Which gets you to the next piece of this, because testing pricing is less straightforward than it sounds.