A B2B SaaS pricing page draws 16.5% of all website traffic, 13 times the visitors of the demo page, and turns 3.8% of them into a form, according to HockeyStack's December 2025 study of 80 companies and more than 31 million visitors. Good pricing page design decides whether those visitors get their answer on the page or book a call to ask for it, and the same study shows the trade: pages that hide prices collect 4.6% of visitors as forms, pages that publish them 2.8%, but the published-price forms become pipeline 17.50% of the time against 10.31%.
This guide covers the design side of that trade. On 7 November 2026 we read 15 live pricing pages and recorded how each lays out plans, badges one, handles billing periods, prices the variable, compares features and routes buyers to sales. The six patterns below come from that read, from HockeyStack and Seeto's data, and from Nielsen Norman Group's guidance. Our pricing page statistics post holds the wider numbers; this one is the layout.
What a pricing page has to do before anyone talks to sales
Nielsen Norman Group's B2B research puts price at the top of the information business buyers need online, and reports that participants went to competitors' sites when a site did not show it. For complex pricing its advice is typical scenarios or a range, because the calculators in its studies were complex, time-consuming and error prone. That guidance is from 2013 and the behaviour holds: in HockeyStack's 2025 data, 62.5% of the 80 companies publish prices, and visitors on transparent sites view 4.26 pages per session against 2.57.
| Measure | Hidden prices | Published prices |
|---|---|---|
| Pricing page visitors who submit a form | 4.6% | 2.8% |
| Form submissions that become pipeline | 10.31% | 17.50% |
| Pricing page bounce rate | 35% | 42% |
| Pages per session, whole site | 2.57 | 4.26 |
| Time on site | 2 min 31 s | 3 min 25 s |
Source: HockeyStack, The State of Pricing, Demo, and Case Study Pages, 80 B2B SaaS companies, December 2025
The first column converts curiosity into calls. The second converts fewer visitors and wastes less of the sales team's week. A call that exists only to reveal a number is the cost the patterns below remove.
15 pricing page design examples, read on 7 November 2026
We fetched each page with a desktop browser request and read the served text, so the cards, badges, toggles and tables below are what a first-time visitor sees. Two more pages, Gong and HockeyStack, show no prices and route every visitor to a form; they are the counterexample, not part of the 15.
| Page | Plans | Highlighted plan and label | Billing control | Stated saving | Top tier | Table and FAQ |
|---|---|---|---|---|---|---|
| Slack | 4 | Business+, "Best value" | Both prices printed | None | Contact sales | Both |
| Notion | 4 | Business, "Recommended" | Monthly / yearly toggle | Up to 20% | Request a demo | Both |
| Linear | 4 | None | Yearly price shown | None | Contact sales | Table only |
| Figma | 4 | None | Toggle on one plan | None | Contact sales | Both |
| Zapier | 4 | None | Task slider plus toggle | 33% | Contact sales | Both |
| Basecamp | 5 | Pro Plus, "The sweet spot" | None, flat monthly | None | Priced, $199 a month | Table only |
| Loom | 4 | Business + AI, "Most popular" | Team size plus toggle | Up to 17% | Contact sales | Both |
| Calendly | 4 | Teams, "Popular plan" | Toggle | Up to 20% | From $15k a year, 50 seats | Both |
| Asana | 4 | None | Toggle | Up to 18% | Contact sales | Both |
| Vercel | 3 | Pro, "Popular" | None, usage credit | None | Get a demo | Both |
| Supabase | 4 | Pro, "Most popular" | None, overage rates | None | Contact us | Both, plus calculator |
| Attio | 4 | Pro, "Popular" | Toggle | 20% | Talk to sales | Both |
| Postmark | 4 | Pro, "Most popular" | Volume slider | None | Contact form | FAQ only |
| Sanity | 3 | Growth, "We recommend" | None | None | Contact sales | Both |
| Ahrefs | 4 | None | Annual option | Up to 17% | Talk to sales | Both |
Source: Octo Agency read of 15 pricing pages, 7 November 2026. Raw file: research/OB-0103-pricing-pages-2026-11-07.csv
Plan count. Twelve of 15 show four plans, almost always free, two paid, enterprise. Vercel and Sanity show three. Basecamp shows five flat-priced packages and no enterprise tier, which fits a product sold to small teams rather than procurement.
Highlighting. Ten pages badge one plan. The wording varies, from "Most popular" on Loom, Supabase and Postmark to "We recommend" on Sanity and "The sweet spot" on Basecamp, but the position barely does: on eight of the ten the badge sits on the plan directly below the top tier. Supabase and Postmark are the two exceptions, with the badge one step lower.
Billing. Seven pages state a percentage saving for annual billing: 17%, 17%, 18%, 20%, 20%, 20% and 33%, a median of 20%. The usage-priced pages, Vercel, Supabase and Postmark, have no annual control at all, because the variable is volume, not months.
Six pricing page design patterns that lower sales calls
| Pattern | Evidence | Pages that do it |
|---|---|---|
| 1. Publish the number, then the floor | Transparent-price forms reach pipeline at 17.50% against 10.31% (HockeyStack, 2025) | Calendly prints $15k a year and 50 seats; Basecamp prices every tier |
| 2. Four plans, one badge, below the top tier | 10 of 15 badge a plan; 8 of 10 place it one below the top (Octo read, 2026) | Notion, Loom, Calendly, Attio, Sanity |
| 3. One toggle, one stated saving near 20% | 80% of 600 pages discount annual billing by 17% to 20% (Seeto, 2026) | Notion, Calendly, Attio, Asana |
| 4. Price the variable in the card | Hybrid pricing at 37% of companies, up from 25% (Growth Unhinged, 2026) | Supabase, Postmark, Ahrefs, Sanity |
| 5. A comparison table of five or fewer, differences first | NN/g static-table limit of 5 items; let users hide identical rows (NN/g, 2024) | Linear, Zapier, Attio |
| 6. Sales as a route, not a wall | Hidden-price pages collect 4.6% of visitors as forms but only 10.31% reach pipeline (HockeyStack, 2025); 14 of 15 pages pair a priced plan with a sales CTA | Vercel, Figma, Slack |
Sources: HockeyStack, December 2025; Seeto, April 2026; Growth Unhinged, May 2026; Nielsen Norman Group, February 2024; Octo Agency read, 7 November 2026
1. Publish the number, then publish the floor. The enterprise card is where most pages stop being transparent: Seeto found only 36% of 600 pages publishing an enterprise price. Calendly's card reads "Starts at $15k/yr" and "Starts at 50 seats", which tells a 12-person team not to book the call and a 200-seat buyer that the budget line is realistic. That one line replaces two kinds of wasted meeting.
2. Four plans, one badge, one step below the top. The four-column layout is the convention because it reads left to right as a ladder: start free, pay for the team, pay for control, talk about scale. The badge sits one step below the top in eight of ten pages for a reason: it marks the last plan a buyer can purchase without a call. Keep every card the same height so the badge, not the card size, carries the emphasis.
3. One toggle, one stated saving. NN/g's toggle guidance is that a switch suits two mutually exclusive states with a default and an immediate effect, which is exactly monthly against annual. Three things make the control work on the pages we read: the saving is written next to the switch ("Save up to 20%" on Notion and Calendly), the default is annual so the first price seen is the low one, and the per-period wording changes with it ("per user/month, billed annually" on Attio). Zapier's 33% is the outlier; Seeto's data says buyers expect 17% to 20%, so a larger figure reads as a monthly price raised to fund it.
4. Price the variable in the card. The newest pages put the overage rate where the price used to be alone. Supabase's Pro card reads "From $25 / month", then "100,000 monthly active users then $0.00325 per MAU" and "250 GB egress then $0.09 per GB". Postmark prints "Extra emails @ $1.30 / 1,000" under each plan and moves the headline price with a volume slider. Ahrefs prints "Add 2 more users at $40/mo each" inside the Lite card. Growth Unhinged's survey of 230 companies found hybrid pricing at 37%, so more buyers arrive with a usage question, and a card that answers it is a call that does not happen.
5. A comparison table that shows differences first. Fourteen of the 15 pages carry a feature comparison under the cards, and most are long: Linear's runs to dozens of rows across four columns. NN/g's rule for a static table is five items or fewer, meaningful attributes, short cell text, and a control that hides rows where every plan matches. Baymard's plan-matrix research adds the failure mode: a user who cannot quickly find a feature critical to their business concludes that no plan offers it. On mobile, NN/g notes most dynamic tables hold three to four items; a four-plan table at 375 pixels needs a plan switcher or a two-column view, not a horizontal scroll.
6. Sales as a route, not a wall. All 15 pages have a self-serve button on at least one priced plan, and 14 also offer a sales route. The design question is placement. Vercel puts "Start a free trial" on Pro and "Get a demo" on Enterprise, so the route matches the tier. Figma adds "Want more information? Contact sales" under the Organization card, between self-serve and enterprise, for the buyer who is unsure which side of the line they are on. Gong and HockeyStack, by contrast, show no price and a form, which HockeyStack's own data says will collect more forms and qualify fewer of them. Our demo request page post covers what to put on the page that form leads to.
Seat, usage and hybrid pricing need different layouts
| Model | Share of pages (Seeto, 600 pages) | What the card needs | Example |
|---|---|---|---|
| Per seat | 58% | Seat rule, per-period wording, annual toggle | Notion, Attio, Calendly |
| Usage-based | About 22% | Included volume, overage rate, a slider or calculator | Postmark, Supabase |
| Hybrid, base plus usage | About 15% | Base price, included credit, metered rate | Vercel, Figma |
| Flat per package | Not reported | User cap and limits per package, no toggle | Basecamp |
Sources: Seeto, 600 SaaS pricing pages, March to April 2026; Octo Agency read, 7 November 2026
Seeto's three-year comparison shows per-seat falling from about 70% to 58%, "Contact Sales" buttons roughly doubling, and the median page carrying two add-ons, most of them AI features. Growth Unhinged puts AI credit adoption at 29%, with a further 33% planning it within a year. Both trends mean more numbers per card and more rows in the table. The layout answer is a second-level control (Figma's seat-type tabs, Attio's credit-pack table, Supabase's compute section) below the cards, so the first screen stays at one price per plan.
Pricing page design ideas worth testing, and what to measure
A pricing page carries enough traffic to test, which most B2B pages do not. Plan for one win in eight and pick the tests with the largest expected effect; our A/B testing statistics post has the win-rate data. Growth Unhinged found that 42% of companies in its pricing index changed prices in the first three quarters of 2024, with an average rise of 20%, and that three in four changed pricing or packaging in the year to May 2026.
| Test | Change | Primary metric | Watch for |
|---|---|---|---|
| Badge position | Move "Most popular" one tier down or up | Clicks on the badged plan | Average contract value falling |
| Enterprise floor | Add "Starts at" price and seat minimum to the top card | Sales-qualified share of demo forms | Fewer forms, which is expected |
| Default billing period | Annual default against monthly default | Paid sign-ups and annual share | Monthly-first buyers dropping |
| Overage in the card | Print the metered rate under the base price | Pricing page to sign-up rate | Support tickets about billing |
| Differences-only table | Collapse rows where every plan matches | Scroll depth to the FAQ, plan clicks | Buyers missing an included feature |
| Sales CTA placement | Sales link under the second-highest tier against enterprise only | Demo forms from priced plans | Self-serve sign-ups falling |
Source: Octo Agency test order, built on HockeyStack 2025, Seeto 2026 and NN/g guidance
On a pricing page build we would wire the billing toggle, every plan button, trial starts, the table's expand control and the sales form as separate events from day one, so each test above has its metric before the first variant ships. Otherwise the only number is form count, which HockeyStack's data shows is the metric that hiding prices improves.
What this means for your pricing page rebuild
Design the page for the visitor who will never book a call: four plans, one defended badge one step below the top, a single toggle with the saving written beside it, the variable priced inside the card, a comparison table that hides what is identical, and a sales route beside the self-serve button rather than in front of it. Then print the enterprise floor, and accept that forms will fall while pipeline per form rises. We build marketing sites in Webflow and have a bias towards it, but the patterns are platform-neutral; our marketing website service page explains how we scope a page like this.


