Audit a SaaS or product pricing page for clarity and conversion, then separate what you can fix now, what is a pricing-strategy question, what is a testable presentation hypothesis, and what still needs a manual check. This is a decision audit, not a scored conversion prediction.
No login. Autosave in your browser. The page URL is context only and is never fetched or crawled.
The audit adapts to your pricing model and purchase motion. It does not assume self-serve tiers.
Architecture facts
Enter what the page shows. A price with no CTA, or a plan with no audience, becomes a concrete finding.
If the three numbers are present and inconsistent, the audit reports a correctness blocker, not a test.
Tick anything the team has not settled. These become upstream questions, not page defects.
Optional. If you enter funnel metrics they are treated as your reported evidence, not measured truth and never as causation.
Items that do not apply to your context are excluded. “Not sure” becomes a verification task, never a pass.
2 item(s) hidden as not applicable to this context.
3 item(s) hidden as not applicable to this context.
3 item(s) hidden as not applicable to this context.
2 item(s) hidden as not applicable to this context.
2 item(s) hidden as not applicable to this context.
Complete the context and checklist, then select “Run audit”.
Pricing strategy decides the price level, the value metric, packaging and segmentation. A pricing page only communicates those decisions well or badly. Many pages fail before the first pixel: the team has not chosen what customers pay for or what changes between plans, so no amount of copywriting will make the page coherent.
This tool audits the communication layer against the architecture you already chose. When it detects that the architecture itself is unresolved, it records an upstream pricing-strategy question instead of diagnosing the page. Setting prices, choosing a value metric and willingness-to-pay research belong to pricing strategy owners, see the pricing experiments guide.
The audit covers six groups: plan comprehension (who each plan is for, what changes between plans, whether a buyer can estimate their cost), price and value communication (the pricing unit, scaling, transparency where sales-led, annual versus monthly), decision path and CTA (an explicit action per plan, what happens after the click, a recommendation or rule), objections and risk (commitment, cancellation, refunds, overages, FAQ), trust and proof (social proof, security and procurement for business buyers, support expectations), and experience and access (narrow-screen comparison, keyboard and screen-reader usability, unambiguous prices).
A self-serve B2C page is judged on a clear unit, simple plan choice, an explicit trial or free path and a matching CTA. A sales-led B2B page is judged on transparency rather than a public number: does it explain what is knowable now, who the sales path is for, and what a buyer gets before contacting sales? A hybrid page must serve both without pretending the enterprise buyer will self-serve. The audit changes its checks with the purchase motion instead of applying one template to all three.
Usage-based and hybrid pages often have no classic good/better/best tiers, so “plan count” rules fail them. Instead the audit asks whether cost is predictable as consumption grows, whether overage rates and thresholds are shown before purchase, and whether add-ons and credits are disclosed. A coherent hybrid page states the platform fee, the usage unit and the overage behaviour together.
Correctness defects are not experiments. A contradictory annual saving, a plan with a price and no CTA, a placeholder price, or a missing commitment term should simply be fixed. A present-but-unclear element, a plan recommendation that is visually ambiguous, CTA wording, proof placement, annual-savings framing, is a genuine hypothesis with more than one plausible option, so it becomes an experiment candidate for the A/B Test Plan Generator. The audit keeps those categories separate instead of returning “five tests”.
A hybrid usage page lists Starter (free + usage), Scale ($99/mo + usage) and Enterprise (custom), and its banner claims “Save 25% on annual”. The monthly/annual prices imply about 20%. The audit reports a blocker: the displayed saving contradicts the prices, a fix, not a test. A missing overage-rate disclosure and an absent cancellation summary are further fixes. A vague scaling description is present-but-unclear, so it becomes an experiment candidate rather than a defect. The comparison table has not been checked at 360px, so that is a manual verification item. Loading the “Hybrid usage pricing” example reproduces exactly this shape. All example data is synthetic and illustrative, not a benchmark.
The audit is deterministic and browser-based. It does not fetch the URL, read analytics, or predict conversion. From pasted copy alone it cannot prove:
It also cannot tell you the right price, value metric or packaging. Those are decisions, not observations.
Symptom: The page cannot be audited because the pricing model is still undecided.
Cause: Value metric, packaging or price level has not been chosen, so the page has nothing coherent to communicate.
Fix: Route value metric, packaging and price level to pricing research first; treat the page as a downstream communication layer.
Symptom: A contradictory number is treated as a design problem.
Cause: The claimed annual saving does not match the monthly and annual prices, but the team plans an A/B test.
Fix: Fix contradictory prices immediately; they are correctness defects, not hypotheses.
Symptom: Sales-assisted pricing is penalised for not showing a number.
Cause: The audit assumed self-serve transparency and treated every hidden price as a defect.
Fix: Judge clarity instead: does the page explain what is knowable now, who the sales path is for, and what the buyer gets?
Symptom: Usage or hybrid pricing is forced into a three-tier template.
Cause: The audit assumed fixed plans and failed a model that has no classic good/better/best tiers.
Fix: Model the architecture first (usage, credits, hybrid) and only then ask whether scaling and overages are clear.
Symptom: Visual and accessibility findings are asserted from pasted copy.
Cause: The report claimed to have inspected layout, mobile behaviour or screen-reader output it never saw.
Fix: Separate what copy can prove from what needs a live-page review; route the rest to manual verification.
Symptom: A single score becomes the goal.
Cause: A 0–100 number implies a conversion prediction that no checklist can support.
Fix: Drop the score; use the fix / strategy / experiment / verify register to decide what to do next.
It audits the page: whether it clearly communicates the pricing architecture you already chose. It does not set the price, choose the value metric or design packaging. When the page cannot communicate pricing because the strategy is unresolved, the audit routes that question upstream instead of blaming the copy.
A numeric score in this space implies a conversion prediction that no checklist can support. Competitor scanners often pair a score with benchmark claims that cannot be verified. This tool shows a grouped, scoreless audit plus a fix / strategy / experiment / verify register, which is more useful for deciding what to do next.
No. The URL is stored as context only and is never fetched. Findings are based on the pricing model, plan rows and checklist states you provide. Because of that, the tool is explicit about what it cannot see rather than implying live page analysis.
A fix is a page defect that is simply wrong (contradictory price, broken CTA, missing commitment term). A strategy question is an unmade pricing decision that belongs upstream. An experiment is a present-but-unclear element with multiple plausible options. A verification item is something the tool cannot inspect, such as mobile layout or keyboard order.
Sales-assisted pricing is not treated as a defect. When the purchase motion is sales-led, the audit asks whether the page explains what is knowable now, who the sales path suits, and what the buyer gets. Hidden exact pricing is only a problem when the page does not explain it.
Yes. The audit models flat, tiered, per-seat, usage, hybrid, freemium, credits and custom-quote architectures. Usage and hybrid models unlock scaling, overage and add-on checks instead of being forced into a fixed three-tier template.
If you only paste the page copy, the live page is not visible, so visual items such as layout, comparison behaviour and accessibility are listed under manual verification rather than as defects. Page-reviewed mode lets you record them as observed findings.
Each experiment candidate comes with the observed issue, hypothesis, candidate change, expected mechanism, primary metric, guardrail and evidence needed. Copy it into the A/B Test Plan Generator, which owns sample size, duration and the statistical contract. This tool does not calculate statistics.
Three synthetic examples cover a self-serve B2C subscription, a B2B sales-led page and a hybrid usage page with a contradictory annual saving. They are illustrative, not benchmarks. You can export Markdown, JSON (lossless) and CSV (findings register), print the report, and share a link.
Everything runs in your browser and autosaves locally. Share links compress the audit into the URL, which means the data travels in the link, do not share it publicly if it contains unreleased prices or commercial terms. No pricing data is sent to analytics.
Go deeper on pricing experiments, trial versus freemium decisions, and measurement.
Build the pricing communication and experimentation skills to act on the audit, not just produce a report.
Last updated: 2026-09-24