How many pricing plans should you offer? One for sorrow, three for joy

· Founder of Magpie AI

Published · 8 min read

In short

Most products do well with two to four plans, and three is a sound default, as long as each one suits a buyer you can name and differs from the next in a way that buyer can see. Mark the plan you'd honestly recommend. Skip decoy plans built only to make another look cheap, and say plainly what yearly billing saves.

Contents

Three is a fine place to start. Honestly, though, the number matters less than you'd think, and I'd take five well-told plans over three that blur into each other. What counts is that each plan belongs to a buyer you could describe in a sentence, and that a visitor can see within a few seconds what moving up a plan gets them.

If you're still deciding whether to put prices on your site at all, showing your prices comes first. This one picks up once the plans are definitely going on your site.

How many plans to offer#

Two to four covers most products. One works when every buyer needs the same thing. A fifth earns its place only if there's a fifth kind of buyer, say an agency running twenty clients where everyone else runs one.

I'd write down who each plan is for before writing a single price. "A café with one site and up to 15 staff" is a buyer, whereas "Pro" is just a word. If two plans end up describing the same buyer, they're two eggs in one nest and one of them should go.

Here's a made-up shift rota tool for cafés at staffloop.example, worked through:

PlanWho it is forPrice
SoloOne café, up to 15 staff£19 a month
GroupUp to 5 sites, shared staff£59 a month
Chain6 sites or more, payroll export£149 a month

A café owner can find themselves in that table without reading a single feature, because every row is a size of business they'd recognise.

What the famous jam study really found#

The jam study is a lot smaller than its fame. In 2000 Sheena Iyengar and Mark Lepper set up a tasting booth in a grocery store on two Saturdays, showing 24 jams for some hours and 6 for others. More people stopped at the big display, 145 of 242 passers-by against 104 of 260. Of those who stopped, though, nearly 30% bought jam after seeing six, against 3% after seeing 24. In actual people that's 31 buyers against 4.

Striking, and later work didn't bear it out as a rule. Benjamin Scheibehenne and colleagues pooled 50 experiments with 5,036 people in 2010 and found a mean effect of "virtually zero", with wide swings between studies.

Then in 2015 Alexander Chernev and colleagues ran a second meta-analysis, over 99 observations and 7,202 people, and did find an effect once they sorted the studies by four things. Those were how complex the set of options was, how hard the task, how unsure people were of what they wanted, and whether they were trying to decide with as little effort as possible. It's the paper I'd read if you're building a pricing table, because all four describe the person on your pricing page pretty much exactly. They don't know your product well yet. They're comparing columns of limits, and most of them would like to be done quickly. So I'd rarely cut plans, and I'd spend the time making them easy to tell apart instead.

Why buyers pick the middle plan#

Itamar Simonson called it the compromise effect, back in 1989. In his studies a brand gained share when it became the middle option in a set, and more so when people expected to explain their choice to someone else. That last bit fits business software rather well, since a buyer who has to justify a purchase to their manager can always defend the middle plan.

Buyers tend to perch on the middle, so put the plan you most want to sell there, and make it the one you'd honestly recommend to a typical customer. Then mark it. A badge like "Most cafés pick this" works much like a default, and defaults pull hard. Jon Jachimowicz and colleagues' 2019 meta-analysis, across 58 studies with 73,675 people, put the effect size at 0.68 and found it stronger in consumer choices and where people read the default as a recommendation. Which means the badge has to be true. If it says most people pick it, most people should.

Simonson's studies had people choosing between described products in a study setting, and I couldn't find a published field test of three-plan pricing pages anywhere. I'd treat the middle as a sound habit, then, and stop short of calling it a law.

Do decoy plans work?#

Sometimes, in studies where every detail is a number, and that's about as far as the evidence goes. A decoy is a plan built to lose, worse than one of the others and barely cheaper, so that the other one looks like a bargain beside it. Joel Huber, John Payne and Christopher Puto first showed the effect in 1982, and in their studies the decoy itself was picked about 2% of the time.

Most pricing advice reaches for Dan Ariely's example instead. In his 2008 TED talk he describes an old offer from The Economist (online for $59, print for $125, or print and online for $125), which he gave to 100 MIT students before giving a version without the print-only option to another 100, and the most popular choice became the least popular. Those were students picking on paper with no money spent. A classroom demonstration, in other words, and a long way from a test on a live site.

Shane Frederick, Leonard Lee and Ernest Baskin looked harder in 2014. They couldn't replicate any of the five earlier studies they found that used realistic products, and across 19 studies of their own, where the options had to be experienced or looked at rather than read off as numbers, they found no effect. Shown as numbers, the same products brought it back. The original authors replied that same year that the effect holds when the original conditions are repeated, and fades when buyers already know what they want.

A pricing table is mostly numbers, so a decoy might shift a few choices on yours. I'd still leave it out. Someone will buy the plan nobody should buy, by mistake, and then you've got a customer who feels tricked writing in to support.

How to lay out the table#

Nielsen Norman Group's 2024 guidance on comparison tables suggests them for up to five items, with column headers that stay on screen as you scroll (people "will easily forget which column is for which product") and a way to hide the rows where every plan is the same.

The rows are where it tends to go wrong. Baymard Institute's 2021 study of software sites found plan tables listing "on average 40 distinct features". Baymard's fixes are grouping features under labelled subheadings and putting the groups most buyers need first.

I'd do five things:

  • Put the plan names, prices and buttons in a header that stays visible as the table scrolls
  • Write each limit in words ("up to 15 staff", "5 sites"), never a tick that leaves the buyer to guess the number
  • Lead with the two or three rows that really separate the plans, then group the rest
  • Order the plans from the smallest buyer to the largest, so a visitor finds their own size by reading along (I found no good test of left-to-right order, so this is only what reads naturally)
  • Give each plan's button words that say what happens next, as with any call to action button

On a phone, five columns of ticks won't fit. Nielsen Norman Group suggests turning the columns into tabs there, so a visitor reads one plan at a time and switches between them.

Monthly or yearly billing#

Whichever side a monthly and yearly switch opens on is a default, so everything above about defaults applies to it as well. It's worth having when a real share of your buyers would pay yearly. Open it on the billing you'd recommend to most of them, and don't pick yearly just because it makes the monthly figure look smaller.

Say what yearly saves, in money. The UK's Competition and Markets Authority, in its November 2025 guidance on price transparency, says that where paying monthly and paying yearly cost different amounts, a trader may mislead buyers unless it makes that clear, for example by showing each option's total or by saying "save £X". For a yearly plan paid up front it wants the full price in the offer itself. On staffloop.example that'd be "£190 a year, save £38" beside "£19 a month", and never a lone "£15.83 a month". The guidance covers selling to consumers (and I'm not a lawyer, so take this as a reading of it rather than legal advice), though I doubt a business buyer hit by a surprise £190 charge feels much different about it.

I'm less sure than most pricing advice that a yearly plan keeps customers longer. The one field study I found comparing the two was at three US gyms, where Stefano DellaVigna and Ulrike Malmendier followed 7,752 members over three years, and members on a monthly contract were 17% more likely to stay enrolled beyond a year than those on a yearly one (which ran out unless renewed). The authors put it down to overconfidence, people expecting more of their future selves than they managed. A gym's a long way from software, and a yearly software plan usually renews by itself (as most free trials that take a card do), so it proves little either way. I'd read it as a reminder that a yearly discount buys you cash up front, and not necessarily loyalty.

What a pricing check looks at#

Magpie's full report reads a site's plans beside what its rivals charge, found by web search. It looks for limits stated in words, a marked usual choice and buttons that say what each plan starts, and it compares like with like, so a per-seat price is set against a flat one at one stated team size, and a plan billed monthly or yearly counts as one plan.

Questions people ask#

Is an Enterprise column worth adding?#

Only when large buyers really do need a different contract, their own security review, say, or invoicing terms your other plans don't offer. If you add one, I'd still give it a starting price or a typical one (for the reasons in showing your prices), and I'd keep it a bit out of the way, maybe as a line under the table rather than a fourth column, because the people choosing between your main plans are most of your buyers and the Enterprise column mostly just sits there taking up a quarter of the width. That's a judgement call, mind. If big contracts are where most of your money comes from, give it the full column.

How big should the yearly discount be?#

I found no research figure to copy. The gyms in that study gave two months free, about 17% off a year of monthly fees, which is a common enough offer. Whatever you choose, say the saving in money and make sure it's true.

Sources

  1. Iyengar and Lepper, When choice is demotivating (Journal of Personality and Social Psychology, 2000) (opens in a new tab)
  2. Scheibehenne, Greifeneder and Todd, Can there ever be too many options? (Journal of Consumer Research, 2010) (opens in a new tab)
  3. Chernev, Böckenholt and Goodman, Choice overload: a conceptual review and meta-analysis (Journal of Consumer Psychology, 2015) (opens in a new tab)
  4. Simonson, Choice based on reasons: the case of attraction and compromise effects (Journal of Consumer Research, 1989) (opens in a new tab)
  5. Jachimowicz, Duncan, Weber and Johnson, When and why defaults influence decisions (Behavioural Public Policy, 2019) (opens in a new tab)
  6. Dan Ariely, Are we in control of our own decisions? (TED, 2008) (opens in a new tab)
  7. Frederick, Lee and Baskin, The limits of attraction (Journal of Marketing Research, 2014) (opens in a new tab)
  8. Huber, Payne and Puto, Let's be honest about the attraction effect (Journal of Marketing Research, 2014) (opens in a new tab)
  9. Nielsen Norman Group, Comparison Tables for Products, Services, and Features (Kate Moran and Taylor Dykes, 2024) (opens in a new tab)
  10. Baymard Institute, SaaS UX: 5 Ways to Improve the Scannability of the Plan Matrix (Sally Collins, 2021) (opens in a new tab)
  11. Competition and Markets Authority, Unfair commercial practices: price transparency, CMA209 (2025) (opens in a new tab)
  12. DellaVigna and Malmendier, Paying not to go to the gym (American Economic Review, 2006) (opens in a new tab)

About the author

· Founder of Magpie AI

Alex founded Magpie AI and writes about what turns a landing page's visitors into customers, drawing on what its audits find on real sites, from the first screen to the checkout

Related articles

Fix what loses you customers.Fix what loses you customers. Then bring more of them in

  • Free score, no sign‑upNo sign‑up
  • No credit card neededNo card needed