Two products can have the same cost, the same quality and the same buyer, and sell at wildly different prices, purely because of how the choice was laid out. The number on the tag is the visible part. The mechanism around it, how many options sit beside it, what they cost, what gets compared to what, is where most of the work happens.

The quick version

  • Five mechanisms, one job: tiers, bundling, anchoring, the decoy effect and versioning all exist to help a customer locate value by comparison, because almost no one can judge a price in a vacuum.
  • They are structure, not discounting. Each changes what gets compared to what, which moves choices without you cutting the headline price.
  • The evidence is real and old. The core findings come from peer-reviewed work, anchoring (Tversky & Kahneman, 1974), bundling (Adams & Yellen, 1976), the decoy (Ariely), versioning (Shapiro & Varian, 1998), not pricing-blog folklore.
  • The honest caveat: these are levers on perception. They lift conversion at the margin; they do not rescue a product people do not want, and a manipulative decoy can cost you the trust that makes the next sale.

The idea in depth

Start with the finding that underwrites all five mechanisms: people are poor at judging absolute value and good at judging relative value. That asymmetry is not a marketing slogan, it is one of the most replicated results in behavioural science, and it is why a price almost never works alone.

Anchoring: the first number sets the frame

In their 1974 Science paper, Amos Tversky and Daniel Kahneman showed that when people estimate an uncertain quantity, they start from whatever value is in front of them and adjust, usually not enough. In one demonstration, asking people to estimate 8 × 7 × 6 × 5 × 4 × 3 × 2 × 1 produced a median guess of 2,250; presenting the same multiplication in ascending order, 1 × 2 × … × 8, produced a median of just 512. The opening digits anchored the estimate. (The correct answer, for the record, is 40,320, both groups were wildly low, but the order of the numbers moved them by a factor of four.)

So the move is: control the first number a buyer sees. Lead with the premium tier, an enterprise quote, or a "usually $X" reference before the offer price, so every later number is judged against a high anchor rather than against zero. The same product feels cheaper next to $200 than next to nothing.

Where it breaks down: anchors that are obviously arbitrary or insulting ("was $999, now $99" on a thing that was never $999) are increasingly seen through, and in several jurisdictions fake "was" prices are unlawful. An anchor has to be plausible to work.

Tiers and the decoy: shaping the choice between options

Once there is more than one option, the arrangement starts doing the selling. Two mechanisms govern this. The first is good-better-best tiering: offering a stripped-down option, a standard one, and a loaded premium. Writing in Harvard Business Review (2018), pricing strategist Rafi Mohammed argues a third tier rarely just adds a price point, the cheap "Good" option recruits price-sensitive buyers, the "Best" captures those who want everything, and crucially the premium makes the middle look reasonable by comparison. There is a perceptual tailwind here too: Valenzuela and Raghubir (Journal of Consumer Psychology, 2009) found a "center-stage effect", across five experiments, people assumed the middle option in a row was the most popular and chose it more often, simply because of its position.

The second is the decoy effect, popularised by behavioural economist Dan Ariely in Predictably Irrational. He tested a real Economist subscription layout on students: web-only at $59, print-only at $125, and print-plus-web also at $125. With the seemingly pointless print-only option present, 84% took the $125 bundle and 16% took web-only. Remove the decoy, and the result flipped, 68% chose the cheap web-only option. The print-only tier sold almost nothing itself; its job was to make the bundle look like a steal.

So the move is: design three tiers, not two, and put the option you want chosen in the middle. Use a deliberately weaker "asymmetrically dominated" option (more expensive for less, or the same price for less) to make your target tier the obvious winner, that is the decoy working for you rather than by accident.

Where it breaks down: a decoy a customer notices is a decoy reads as a trick, and erodes trust. The center-stage and tiering effects are robust on simple, comparable options; pile on too many tiers and you trigger choice overload, where the buyer defers or walks.

Bundling and versioning: packaging value to capture more of it

The last pair are about what you sell together and how many grades of it you make. Bundling has a precise economic logic, set out by William Adams and Janet Yellen in the Quarterly Journal of Economics (1976): when different customers value different parts of your offer, selling the parts as one package smooths out those differences and lets you capture value you would lose selling à la carte. A cinema-goer who loves the film but is indifferent to popcorn, sitting beside someone with the reverse taste, will each pay more for a combo than either would pay for the item they care less about on its own.

Versioning is the information-age cousin. Carl Shapiro and Hal Varian, in Information Rules and their HBR article "Versioning: The Smart Way to Sell Information" (Nov–Dec 1998), pointed out that for digital goods, software, media, data, the first copy is expensive and every copy after is nearly free. So instead of one price, you build the high-end version and then deliberately subtract from it (fewer seats, a delay, a watermark, no API) to create cheaper versions that let price-sensitive buyers self-select without cannibalising the buyers who will pay full freight.

So the move is: bundle where your customers' tastes differ (it homogenises willingness-to-pay), and version where you can cheaply create a "less" of your best product. In both cases the customer sorts themselves into the price they're willing to pay, you do not have to guess.

Where it breaks down: bundling can backfire when buyers only want one component and resent paying for the rest (the cable-TV complaint), which is why "unbundling" disrupts incumbents. And a version degraded too obviously, crippling the cheap tier out of spite, annoys customers more than it converts them.

flowchart TD
    A(["A price the customer
has to judge"]) --> B(["Can they judge it
in a vacuum?"])
    B -->|"No, almost never"| C(["Give them a comparison"])
    C --> D(["Set the frame:
anchoring"])
    C --> E(["Shape the menu:
tiers + decoy"])
    C --> F(["Package the value:
bundling + versioning"])
    D --> G(["Customer locates value
by comparison, not in isolation"])
    E --> G
    F --> G
All five mechanisms do one job: give the buyer something to compare against. Leaders Loop

A worked example

Imagine a small SaaS tool, call it a scheduling app, currently sold at one flat price of $20 per user per month. Conversion is fine; revenue per customer is flat. The team wants more without cutting price. (All figures below are illustrative, to show the mechanisms, not benchmarks.)

They restructure into three tiers. Starter at $12 (calendar sync only) recruits the price-sensitive. Team at $24, the old product, plus shared availability, sits in the middle, where the center-stage effect quietly favours it. Business at $48 (analytics, SSO, priority support) is the high anchor: most buyers won't take it, but seeing $48 makes $24 read as the sensible choice rather than a price rise from $20.

The $48 tier is partly a versioning play, the same core app with enterprise features added, while Starter is the same app with features subtracted, so customers self-select. The team also bundles a $6 "automations" add-on into Team at no extra charge, cheap to include, and it widens the perceived gap to Starter. The decoy here does honest work: Starter exists partly to make Team look generous, but the features it lacks are ones serious users genuinely need.

The likely outcome, consistent with Mohammed's good-better-best logic: most revenue concentrates in the middle tier, a minority of high-value accounts move to Business and lift average revenue per user, and Starter recovers customers who'd have bounced at $20. No discount was given, the headline price went up. The change was structural, not promotional.

flowchart LR
    S(["Starter, $12
recruits + decoy"]) --- T(["Team, $24
center stage, the target"]) --- B(["Business, $48
high anchor + version"])
    T -.->|"most revenue"| R(["Higher ARPU,
no discount given"])
    B -.->|"lifts average"| R
    S -.->|"recovers walk-aways"| R
The scheduling-app tiers, illustrative figures showing anchor, center-stage and decoy roles at once. Leaders Loop

Frequently asked questions

Isn't this just manipulation?

It can be, and that is the line to watch. There's a real difference between helping a customer compare (a genuine premium tier that some buyers truly want) and tricking them (a fake "was" price, or a decoy that conceals a feature people need). The honest version improves the buyer's ability to choose; the manipulative version exploits an inability to. The first builds repeat business; the second buys one sale and loses the next.

How many tiers should I have?

Three is the workhorse for a reason, it gives you an anchor, a target and a recruiter, and it triggers the center-stage effect without overwhelming the buyer. More than three or four comparable options risks choice overload, where people defer the decision rather than make it. Add tiers only when each one serves a genuinely distinct segment.

When should I bundle versus sell things separately?

Bundle when your customers value the components differently from one another, that's exactly the condition under which Adams and Yellen showed bundling captures more value. Sell separately (or offer "build your own") when most customers want only one part and resent paying for the rest; forcing a bundle there is what makes you vulnerable to an unbundling competitor.

Do these still work if customers know the trick?

Partly. Anchoring and the center-stage effect are surprisingly persistent even when people are told about them, because they operate below deliberate reasoning. But an obvious decoy or a transparently fake anchor backfires once spotted, awareness flips it from a nudge into a reason to distrust you. Build mechanisms that still look fair when explained out loud.

Where does this sit relative to my overall pricing strategy?

These are mechanisms, the structure of the menu, not the strategy that decides your actual price level. Whether you price on value, cost-plus or to penetrate a market is a separate, prior decision (see pricing strategies). The mechanisms here are how you present and package whatever strategy you've chosen, and they only pay off once the underlying unit economics are sound.

Related in the Toolkit

Where to go next