UK SaaS pricing page: structure, examples, and what to charge in 2026

Your pricing page is where intent becomes revenue. Someone clicked "Pricing" because they are weighing up handing over a card. Everything before was interest. Everything after is churn risk. The page itself is the decision.

And yet most UK solo builders treat it like an afterthought. Three tiers pulled from a Lovable template, a GBP symbol slapped on the numbers, done. Then they wonder why a landing page converting at 4% drops to 0.9% at pricing.

This post is about the page itself — the anatomy, the copy patterns, the UK-specific bits like VAT, the mistakes, and how to test it when you are a team of one. If you want the upstream strategy (cost-plus vs value-based, what a pound buys in 2026), read our earlier piece on how to price your SaaS. This one is about the page.

Anatomy of a high-converting pricing page

There are seven elements that earn their place on the page. Everything else is decoration.

1. The headline. One line that tells the buyer what they get for paying you. Not "Simple, transparent pricing." That is filler. Linear opens with "Built for the fastest-moving teams in the world." Ashby leads with "Pricing that scales with you." The job of the headline is positioning, not menu setting.

2. The tier cards. Two or three, max. Four if you genuinely serve distinct segments. Five is a sign you do not know who you are selling to.

3. The featured tier. One card visually promoted — a border, a "Most popular" tag, a different background. You are telling the buyer what to pick. Decision fatigue is real and "most popular" converts measurably better than three equal cards.

4. The feature list. Not everything you do. The 6-10 features that matter per tier, in the same order across tiers, with a clear delta between them. If tier two and tier three have identical lists except one line, the buyer will spot it and pick the cheaper one.

5. The social proof strip. Logos of customers, a review quote, a count of users. Underneath the tiers or between them. Pricing pages without social proof convert roughly 30% worse than pages with it.

6. The FAQ. Real objections. "Can I change plans?" "Do you charge VAT?" "What happens if I go over the limit?" Each one disarms an objection before it kills the sale.

7. The secondary CTA. For every buyer ready to click "Start free trial" there are three who need a call, a demo, or just a human. Give them a path. A Cal.com link under the enterprise tier works. A contact form is the worst version but better than nothing.

Seven elements. If your page also has a video hero, mega-menu, animated comparison, live chat, roadmap preview and newsletter sign-up — you are not building a pricing page, you are building a distraction.

Tier naming — stop calling it Starter / Pro / Enterprise

The default tier names are Starter, Pro, Enterprise. Safe, boring, forgettable. They make you sound like every other tool on the IndieHackers front page.

Tier names have two jobs: signal who the tier is for, and make the upgrade path obvious. "Starter" signals nothing. "Team" tells me it is for teams. Cal.com runs "Free / Teams / Organizations / Enterprise" — every name maps to a buyer. Tines uses "Community / Professional / Enterprise." Linear runs "Free / Basic / Business / Enterprise."

The rule: name tiers after the customer, not the feature set. "Solo" beats "Starter" if you sell to solo operators. "Agency" beats "Business" if you sell to agencies.

Avoid cute names that need a decoder ring. "Hatchling / Sprout / Forest" means nothing on first pass and the buyer has to work out which one they are. Friction kills conversion.

One exception: a genuinely different product class gets a named tier. Notion's "AI" add-on. Stripe's "Premium Support." These are side products, not rungs on the ladder. Name them what they are.

UK specifics that catch people out

Selling from the UK has tax and currency rules that American templates quietly ignore. Get these right and you remove a whole class of support emails.

GBP by default for UK visitors. If you can detect the country (Vercel geolocation, Cloudflare headers, or a crude IP lookup), show pounds first. Showing $49 to a UK buyer forces them to do mental maths and creates a "am I being charged in dollars?" moment. That moment costs you sales.

VAT: inclusive or exclusive, not ambiguous. B2B UK SaaS typically displays ex-VAT — "£49/month + VAT" or "£49/month (excl. VAT)." B2C SaaS should display inc-VAT because the consumer cannot reclaim it. Whichever you pick, say it on the page. "Prices in GBP. VAT added at checkout where applicable." A line that small saves you a dozen emails a week.

Currency switcher. If you sell internationally, a switcher in the top-right or under the tier cards lets buyers see their own currency. Stripe Checkout handles the actual conversion — you just need the display layer. Do not fake it with fixed rates, because when GBP moves 3% you will be quietly under-charging half your market.

Sales tax at checkout, not on the page. Stripe Tax and Paddle calculate VAT, EU reverse-charge, and US sales tax at checkout from the buyer's address. Let them do the work. Keep the page clean.

VAT numbers for B2B. Add a VAT number field at checkout for EU/UK buyers. Paddle handles it natively; Stripe Tax has it. Reverse-charge on valid EU VAT numbers is the default for UK sellers post-Brexit.

Three pricing pages to study (and what to steal)

I have picked three pages I think solo UK builders can learn from. All are UK-accessible, all have been iterated on, all are live right now.

Cal.com. Open-source scheduling, Free / Teams / Organizations / Enterprise, GBP for UK visitors. Steal: the generous "Free forever" tier builds trust before you ask for money. The comparison table below the cards is the cleanest I have seen — dense but scannable, with genuine deltas between tiers. The enterprise card has no price, just a "Contact sales" button and three bullets. No fake £999/mo that everyone knows is a starting point.

Linear. Project management for software teams. Ruthlessly minimal — three tiers, clear feature list, "Most popular" tag on the middle one. Steal: the tier card copy. Each card has a one-line positioning statement ("Ideal for smaller teams getting started with Linear") under the name. It tells the buyer "you are in the right place" before they read a single feature.

Tines. Security automation, Dublin-based. Community (free, self-hosted) / Professional / Enterprise. Steal: the Community tier is not crippled. Real product, not a demo. Engineers install it, run it a month, then push their company to buy Professional. The "Book a demo" CTA under Enterprise is routed to Cal.com — exactly how a solo seller should handle enterprise enquiries.

Three patterns across all three: one featured tier, genuine free or low-tier value, and a comparison table for detail-seekers.

Five common mistakes

1. Anchoring too low. A "Starter" tier at £9/mo sets the anchor for the whole page. If your main tier is £49, buyers are doing "£49 is five times the Starter" maths. Kill the anchor — make the entry tier £29 or skip it and lead with the main plan.

2. Too many tiers. Four is the ceiling. Five is a cry for help. Six-tier pricing pages chasing "every segment" capture none. Decision paralysis is measurable.

3. Hiding the price. "Contact us for pricing" on main tiers signals expensive or inconsistent. Keep it for enterprise only. Everything else has a number.

4. Monthly-only pricing. Offer annual at 15-20% off. Annual billing improves cash flow and roughly halves churn versus monthly.

5. No mention of cancellation. "Cancel any time" on the pricing page converts better than burying it in-product. Buyers are risk-averse — tell them the exit exists before you ask them to enter.

A/B testing pricing without Optimizely

You do not need enterprise testing tools. Here is the 30-day cohort compare method.

Pick one variable — tier names, anchor price, featured tier, or headline. One, not four. Ship version A for 30 days. Track three numbers: pricing page visits, checkout starts, and paid conversions. Ship version B for 30 days. Compare.

If your traffic is under 1,000 pricing page visits per month, 30 days is too short — bump to 60 days, or only test changes expected to produce a 25%+ lift. Small changes on small traffic are untestable.

Plausible and Fathom both handle pricing page goals cleanly. Stripe tells you paid conversions. Your database tells you checkout starts. Do not run two tests at once — you will not know which change caused the lift.

How Claude Code helps build the page

The pricing page is a good Claude Code job because the structure is known and the variables are small. Two prompts I use and adapt:

Prompt 1 — first draft.

Build a pricing page for a UK SaaS product. Product: [your one-liner]. Target buyer: [solo / team / agency]. Three tiers: [name + price + 5-8 features per tier, listed]. Featured tier: [middle]. Show GBP by default. Include a "prices exclude VAT" note. Add a 5-question FAQ covering: plan changes, VAT, usage limits, cancellation, refunds. Output a single React component using Tailwind, no external UI library. Include a currency switcher stub that swaps GBP / USD / EUR in the tier prices.

Run that in Cursor, Replit, or Base44 and you get a working draft in a minute. It will not be production-grade, but the structure will be right. You iterate from there.

Prompt 2 — teardown and rewrite.

Here is my current pricing page: [paste the component or URL]. Critique it as if you were a UK buyer weighing up £29/mo. Identify: unclear tier names, missing trust signals, anchor pricing issues, VAT ambiguity, and CTA weakness. Rewrite the component with fixes applied. Keep it to seven elements: headline, three tier cards, featured tag on middle, feature list, social proof strip, FAQ, secondary CTA.

That second prompt is the one that moves pages from template to performant. Constraints are tight, feedback loop is short — preview, push to staging, iterate.

One note: do not let Claude Code generate tier names or prices. Those are product decisions. Hand it your decisions and let it execute the structure.