We Showed Users All Three Pricing Tiers at Once. They Bought Less
Our pricing page has always bothered people internally. It shows one plan, a single primary card with one price and one button, and tucks the other two tiers behind a drawer you have to open deliberately. Two of our three tiers are effectively hidden.
The objection writes itself: users who would happily pay for the bigger tiers never see them. We're leaving money invisible. Every serious SaaS shows its tiers side by side; there's a reason the three-column pricing page is the industry default.
So we tested it properly. Four variants on a SaaS product I work on:
Control: the single primary card, other tiers behind a drawer.
All tiers: the classic three-column layout, every tier visible with its own price and button, the middle one visually highlighted.
Recommended: the suggested tier rendered large and prominent, the other two shown smaller beside it.
Comparison: a full feature-comparison table across all three tiers, each column with its own CTA.
Every layout that exposed all the tiers lost to the single card. Not one of the three challengers beat control on completed checkouts. The layout that "hides" two-thirds of our pricing outsold the layouts that showed everything.
What the funnel actually showed
The numbers are roughed out, but the shape was consistent across the losing variants: engagement with the pricing page went up. People scrolled more, hovered more, spent longer on the page. The comparison table in particular got real attention.
Then fewer of them bought.
The multi-tier layouts converted attention into deliberation, and deliberation into leaving. The single-card control gave visitors almost nothing to weigh: one price, one decision, yes or no. And "yes or no" beat "which one" decisively.
Choice overload is real, and pricing pages are where it lives
The psychology here is old and well documented. Give people more options and two things happen: the probability that they defer the decision goes up, and their confidence in whatever they pick goes down. The famous jam-study framing gets overused, but the mechanism underneath it kept showing up in our data.
Three tiers on screen means the visitor's job silently changes. On the single card, the question is "is this product worth it?" On the three-column layout, the question becomes "which of these is right for me?", which drags in a dozen sub-questions: What if I outgrow the small one? Am I overpaying on the big one? What does that feature on the middle one actually mean? Every sub-question is a chance to decide later, and later is where conversions go to die.
The comparison table was the purest version of this. It's the layout that respects the buyer most: total transparency, every feature enumerated. It's also the layout that most thoroughly converts a purchase decision into a research project.
The nuance: a wall is not an anchor
Here's where I'd stop you before you delete your tier columns tomorrow.
What we tested, and what lost, was a tier wall: presenting all options as equals and making the user do the choosing. That's different from anchoring: leading with one deliberate option while keeping the others reachable.
Our control was, accidentally, a strong anchor. One card makes the decision binary, and the drawer keeps the other tiers available for the minority who go looking. The losing variants removed the anchor and handed the user a homework assignment.
Which raises the question the losing test can't answer: if one visible option beats three, which option should the one card show? The cheapest tier is the safest ask, but it anchors everyone low. Anchoring on a higher tier, with a graceful one-tap path down, might collect more revenue per visitor without scaring off trials. That's a separate experiment (anchor placement rather than option count), it's running now, and it deserves its own write-up when it concludes. The early lesson stands on its own though: the number of simultaneous options mattered more than which options they were.
Why the industry default exists anyway
If tier walls lose, why does every SaaS pricing page look like one? A few honest hypotheses:
Nobody tested it. The three-column layout is cargo-culted from companies whose context is different (sales-assisted, enterprise-weighted, high brand trust). Defaults propagate because they look professional, not because they won an experiment.
It optimizes for a different funnel. If most of your revenue arrives through a sales call, the pricing page's job is education, not conversion, and a comparison table educates well. Ours is pure self-serve; the page's only job is the checkout.
Segment mix matters. A visitor pool full of committed, tier-aware buyers might genuinely convert better with everything visible. Ours skews first-touch: people still deciding whether to buy at all, for whom extra options are extra doubt.
All three hypotheses point the same direction: the right pricing layout is a function of your funnel, and the only way to know yours is to test against your own traffic.
What we'd tell you to steal
Count decisions, not tiers. Every simultaneously visible option multiplies the cognitive work of the page. Ask what decision you're actually asking the visitor to make, and whether it's the smallest one that gets you paid.
Watch for the engagement head-fake. Our losing variants looked great on time-on-page and interaction metrics. Deliberation reads as engagement right up until you check purchases.
Keep the other tiers reachable, not visible. Hiding tiers behind one deliberate click didn't stop the people who wanted them. Upgrade paths exist after purchase too; you don't have to close the whole sale on one screen.
Separate "how many options" from "which option leads." They're different experiments. Run the count test first; it's the bigger effect.
The three-column pricing page is one of the most confident-looking patterns on the internet. Ours is a single card with a price on it, and it beat three better-looking, more transparent, more industry-standard layouts in a fair fight. Confidence is not evidence. Traffic is.