Use this when
/pricingis a hero sentence and a “Talk to sales” button.- Sales refuses to publish any number, and inbound asks “what does this cost?” on every first call.
- Four editions are on the page and the tenth customer does not exist yet.
- The value metric is obvious internally and invisible in the first five seconds.
Do not use this when
- You have not decided the unit or the first offer. Stay in pricing and packaging.
- Public-procurement or regulated quoting forbids a public number. Write the constraint; this page is not legal advice.
- The work is a multi-product catalog redesign. Position the way you sell first.
- You want to A/B-test plan names before the metric is readable.
A few useful terms
Keep this in mind
Show the unit in five seconds. If a stranger cannot say what they pay for, the rest of the page is decoration. Clever headlines, tabbed product families, and “most popular” on the most expensive tier do not fix an invisible metric.How to do it
Step 1: Make pricing easy to find
/pricing is not a landing page. No newsletter pop-up. CTA above the fold. Restate who this is for in one line (positioning, not a new slogan). Clean structure over a revolutionary layout.
Step 2: Show a price or explain how to get one
Default: yes, show pricing. For a sales-assist motion, a public starting price, a range, or “from $X / [unit]” plus what enterprise adds is still a number. A page that only says “contact us” teaches the buyer you are expensive and opaque. If legal truly forbids a figure, show the metric, the plan jobs, and the questions sales will ask—so the call is not a blank.Step 3: Make the plans easy to compare
- Three plans is the usual; never more than five.
- Summary table: the few differences a buyer uses to self-qualify. Full feature list below the fold.
- Waterfall the features so value visibly rises with price—and the features are actually valuable.
- Suggest a plan with “recommended” or “best value” only if it is true. Do not tag the top tier “most popular” as a lie.
- Add-ons sit beside the table when a capability is used by a small share of a plan (Elena Verna’s operating heuristic: think hard below ~30% use). Do not hide them as fake checkmarks.
Step 4: Give each product a clear entry point
Multi-product tabs are a hub, not a discovery engine. Most people will not click the other tab. Drive the second product with content strategy links and sales motion—not with a tab bar. Salesforce-style box salad is what not to do if you need the page to decide.Step 5: Use visitor behaviour to choose a test
Feature-gated products often see repeat pricing visits before pay. Usage-gated products often convert on the visit that hits the limit. Track unique visitors, not raw pageviews, before you call conversion “low.” If you have not tested the page in a year, you are running a 100% holdout with no log. Tests still need a decision—see experimentation. Do not A/B a strategy question (which metric, which motion) as a headline.Worked example (illustrative)
Sales-assist. Seat-based. Enterprise is SSO, DPA, and a named CSM—not a mystery SKU.Copy: pricing-page brief (fill)
- Value metric (one sentence a stranger can repeat):
- Plans (≤5) and who each is for:
- What is public vs “talk to us,” and why:
- Add-ons (and why they are not stuffed into a tier):
- Primary CTA above the fold:
- Multi-product: single lineup / tabs / separate URLs — and how the others get traffic:
- Unique-visitor conversion we will read (not raw views):
- Last test date / next test allowed to change:
Before you start
- The commercial unit is already decided on pricing and packaging.
- Pricing is in the main nav.
- A stranger can name the unit in five seconds.
- ≤5 plans; summary above, full list below.
- Enterprise path names what is different.
- No pop-up. CTA above the fold.
- “Most popular” is true or absent.
- Self-serve metrics (if any) are unique-visitor based.
Metrics
Elena Verna’s public PLG tape (signups who hit pricing ~25%; checkout-to-order ~50%; reverse-trial free-to-paid ~10–15%) is her self-serve sample. Use it as a method prompt, not as your target. High-ACV sales-assist should not import those percentages.
Common mistakes
- Blank “contact us” as a strategy.
- Four editions for the tenth customer.
- Feature-stuffing so nobody feels they overpay—then they churn because they use one of ten.
- Tabs as the discovery plan for a second product.
- Testing copy before the metric is readable.
- Treating pageviews as conversion.
What to read next
The high-intent form that should skip the leisurely score ladder is demo request. The scan that must still work if they never open/pricing is the homepage. How you will read whether ads or content created this visit is measurement model. The commercial revisit calendar stays on pricing and packaging.
Sources and evidence boundary
This is an owner-maintained operating synthesis.- Show the number; unit in five seconds; three-to-five plans; summary then waterfall; no pop-ups; add-ons when use is minority; tabs do not discover the second product; unique-visitor conversion; contextual in-product upsells. Distilled from Elena Verna, “The DNA of a Great Pricing Page” (2024-10-11), which also points at Emily Kramer / MKT1 as the checklist to follow. Teardowns of Figma, Slack, Miro, Zoom, HubSpot, and Salesforce in that essay are their grades on those pages on that date—not a command to copy a layout.
- Plan count, value metric, and “charge sooner / simpler first offer” stay owned by pricing and packaging.
- This page does not re-gate a Reforge course. PLG benchmarks stay labeled as someone else’s tape.
Copyright © 2026 Ivan Xu. All rights reserved. See the copyright and reuse terms. Canonical source: github.com/weilun88313/B2B-Playbook