Good-better-best pricing: how to design the three-tier ladder

Good-better-best is the most copied pricing structure on the internet, and the most often copied wrong. Three boxes side by side isn't a strategy — it's wallpaper. The actual strategy is in how the tiers anchor each other and which one you secretly want most people to buy.

Why three tiers

Two tiers force a binary yes/no decision. Four or more triggers analysis paralysis — buyers default to not nowwhen they can't compare quickly. Three is the sweet spot: enough choice to feel chosen, few enough to decide in under a minute.

The classic compromise effect from behavioural economics: given three options, buyers gravitate to the middle one because it feels safe — not the cheapest (which signals "you'll regret it"), not the most expensive (which signals "you're being upsold").

The anchor doesn't have to sell

A top tier that few people buy still changes how the middle tier looks: next to it, the middle option reads as the sensible choice. The expensive option doesn't have to sell. It has to exist.

The role of each tier

Good — your entry point. Designed to be good enough to not feel like a trap, limited enough that buyers who care about features upgrade. Goal: convert the price-sensitive segment without devaluing the rest.
Better — your target. This is the tier most of your revenue should come from. Priced so the gap to Good is justified by the visible feature gain.
Best — your anchor. Mostly a pricing device, not a revenue target. Should look obviously premium so Better looks reasonable. A minority of buyers should choose Best; if it becomes your most popular tier, you've under-priced it.

Setting the prices

The 1 : 2 : 4 rule (rough starting point)

One illustrative starting point: Good = X, Better = 2X, Best = 4X. That gap is large enough that Better feels like a meaningful upgrade and Best feels distinctly premium. It is a heuristic, not a law — move the ratios based on your data:

If too many buyers pick Good — narrow the Good→Better gap or remove a key feature from Good.
If too many buyers pick Best — widen the Better→Best gap or add a premium-only feature to Best.
If buyers pick the middle but you wanted more upsell — your Better tier is doing exactly what it should. Add a Best+ tier or accept the win.

The distribution to aim for

In a working ladder, Better carries the majority, Good takes the price-sensitive minority, and Best takes a small share. There is no universal split; track yours over time. If the mix looks wrong, the issue is usually that two tiers are too close on either price or feature set — the ladder has a missing rung.

Feature differentiation that actually works

The hardest part isn't the price — it's deciding what goes in each tier. The principles:

One unmissable reason to upgrade. Each tier jump should have a single headline feature, not a list of 12 minor improvements. "Unlimited users" beats "advanced permissions, custom roles, audit logs" for upgrade clarity.
Quantitative limits, not feature flags."100 / 1,000 / unlimited" upgrades cleanly because buyers can place themselves on the ladder. Feature flags ("Pro only") create resentment when the missing feature feels arbitrary.
Don't gate security or support behind Best.SSO, audit logs, and uptime SLAs in the top tier are industry-standard but increasingly criticised — many buyers see this as a tax on doing things properly.

The decoy effect, used well

If your real goal is to drive buyers to a €99 tier, add a €149 tier that offers only slightly more. The €149 makes the €99 look like the obvious deal. The decoy doesn't need to sell — its job is to shift the choice architecture. (Don't overdo it: buyers smell it and trust suffers.)

The pricing page that supports the ladder

Highlight the middle tier visually — border, badge, "Most popular". This isn't manipulation; it's signalling where your typical customer lands.
Annual toggle on by default if your annual discount is > 15%. Anchors the price to a smaller monthly number.
Feature comparison below, not in the cards. The cards sell the upgrade story; the table answers the "yes-but-does-it-have-X" questions for the long-tail evaluator.
One CTA per card. "Start free trial" or "Contact sales" — never both on the same card. Mixed CTAs confuse buyers about which tier needs a salesperson.

When good-better-best is the wrong choice

Usage-based pricing — if your value scales continuously with usage, a single per-unit price beats three tiers that get gamed.
Enterprise-only products — three packaged tiers create artificial caps that block deals. "Contact us" is fine if your buyer expects a custom quote.
Marketplaces / two-sided platforms — pricing is a take rate, not a tier ladder. Trying to bolt good-better-best on top usually hides the real economics.

For deeper context, see psychological pricing: charm, anchoring and decoys, pricing strategy fundamentals, and price elasticity of demand.

What "good" looks like

Middle tier is your majority. If not, the ladder is mis-rung.
Top tier exists but doesn't carry revenue. Its job is to make Better look smart.
You can re-tier without rewriting the product.Limits, quotas, and seat counts move easily. Feature flags ossify.

Measure whether a price change worked

For online stores, Luska's “Did it work?” view compares sales in the 7 days before and after each approved price change, so you can see whether a new price moved buyers the way you intended.

See price change impact