pricing_page_copy v1.0.0

name: "pricing_page_copy" description: "Write pricing pages that convert without misleading buyers: plan naming and ordering, honest feature comparison tables, anchor and decoy use with limits, total-price disclosure, and FAQ that handles price objections. Triggers on pricing page copy, plan tiers, pricing tables, comparison tables, monthly annual toggle copy, or rewriting a pricing page."

Pricing Page Copy

Purpose

Help a buyer answer three questions fast: which plan fits me, what will I actually pay, and what changes if I pay more. A pricing page fails when it lists plans but leaves those questions to guesswork. This skill gives a page structure, rules for naming and ordering plans, a method for building comparison tables that stay honest, and a test for anchor and decoy tactics so they clarify value instead of manufacturing it.

It covers copy and structure only. It does not set the underlying prices. If the prices, limits, or feature split are undecided, say so and draft the page around the decisions that still need an owner.

Workflow

1. Lock the facts before writing a word

Collect the inputs in writing first. A draft built on assumed facts will read well and be wrong.

• The value metric: what the price scales on, such as seats, projects, usage, locations, or a flat fee. Name one primary metric per plan set.
• The buyer segments each plan serves, in the buyer's own words: solo, small team, department, whole company, or a named job.
• The real limits and overage rules for every plan: what happens at the limit, what extra use costs, and whether anything stops working.
• The full price a buyer pays: base price, mandatory fees, required add-ons, billing period, tax treatment if it is stated, renewal price after any introductory period, and cancellation terms.
• Proof you can show: current customer counts, named customers with permission, review scores with a source and date, or a guarantee you actually honor.
• The one action the page exists to produce: start a trial, buy, book a call, or compare and then choose. Pick one primary action. A second action such as talk to sales is allowed only as a clear secondary path.

If a fact is missing, leave a bracketed placeholder and list it under open risks in the output. Never fill a gap with a plausible number.

2. Name and order the plans

Plan names do work. They tell a buyer where they belong before they read a single feature.

• Name by buyer or job when segments differ in kind, not just size. "For freelancers" and "For agencies" sort a buyer faster than Bronze and Silver, which say nothing about fit.
• Name by scale only when the plans really are the same product at larger volume. In that case keep names short and parallel, and let the limit line carry the difference.
• Avoid cute names that need a legend. If a buyer has to decode the name, the name failed.
• Keep the name, the one-line fit statement, and the price in that order on each card. The fit statement is one sentence: who it is for and the outcome that plan buys.
• Order plans left to right from lowest price to highest for self-serve buyers scanning on a phone or desktop for the first time. Put the recommended plan in the middle when there are three plans, and mark it plainly.
• Consider high-to-low order only for a sales-led or considered purchase where the seller wants the premium scope read first, and record why. Anchoring research shows the first number seen pulls later judgments toward it (see step 5), so order is a choice with an effect, not a default to copy from a competitor.
• Limit the set to the plans a buyer can tell apart in one pass. Research on choice overload found far fewer purchases when shoppers faced 24 jam varieties than when they faced 6, even though the larger display drew more attention (Iyengar and Lepper, 2000). A pricing page is a different setting, but the direction holds: plans a buyer cannot distinguish become a reason to leave. Merge plans whose only difference is a feature few buyers value, and move true edge cases to an add-on or a custom note.

3. Write each plan card to the same contract

Every card carries the same fields, in the same order, so buyers can compare without relearning the layout.

1. Plan name.
2. Fit statement, one sentence.
3. Price, with the billing period attached to the number itself: "$29 per seat per month, billed annually." A price with no period is an incomplete claim.
4. What the price becomes on the other billing period, stated on the card or one tap away with the toggle labeled in words, not just a switch.
5. The 3 to 5 differences that matter most for this plan, written as outcomes or concrete capacities. Limits go in numbers, not in words like "generous" or "advanced."
6. One call to action with the same verb across all self-serve plans. Vary the verb only when the action really differs, such as "Start trial" versus "Talk to sales."
7. Risk reducer under the button where it is true: trial length, no card required, cancel terms, or guarantee period. State the condition, not just the badge.

Price display rules:

• Show the total a buyer commits to, not only a monthly equivalent. If annual billing is shown as a per-month figure, show the annual total beside it.
• Show mandatory fees on the card. Hidden fees and prices were among the most damaging web design mistakes in Nielsen Norman Group usability reporting, and a price a buyer cannot reproduce at checkout costs trust at the exact moment of decision.
• If a price is introductory, name the period and the renewal price in the same block. "First year" with no second-year price is a half claim.
• Use "starting at" only when a buyer can actually buy at that starting price for a real configuration. Name what drives the price up.
• Keep decimal formatting and units consistent across plans. Align numbers so a buyer can subtract one plan from another at a glance.

4. Build the comparison table as a decision tool

The card row handles the quick choice. The full table handles the careful buyer, procurement, and anyone comparing you with a rival. Treat the table as an editorial product: the rows you choose decide what buyers weigh, and an attribute left out of the table barely counts in the decision.

• Group rows under buyer tasks, not internal departments: getting started, day-to-day work, collaboration, admin and security, support. A buyer thinks in tasks.
• Order rows inside each group by how often the difference decides a purchase, gathered from sales calls, support questions, and churn notes. Put the deciding rows first, not the rows engineering is proudest of.
• Lead the table with a short summary version of 5 to 7 deciding rows. Put the full table below it or behind a clearly labeled expand control, and keep the summary honest: do not promote a row to the summary only because you win it.
• Write each row label as the capability itself, in plain words a new buyer uses. Avoid internal feature names with no explanation.
• Fill every cell with a value, a limit, or a clear mark. Use one vocabulary across the whole table: check for included, a number for a limit, and a word like "No" or "Not included" for absent. Never leave a blank cell and expect a buyer to guess whether it means no, zero, or not applicable.
• Define any symbol once, next to the table. Do not use stars, dots, or tiers inside a table unless the scale is printed beside it and applied the same way in every row.
• Mark add-ons as add-ons, with their price or price basis in the cell or in a footnote tied to that row. A check that quietly requires a paid add-on misleads.
• Date the table when feature sets change often, and recheck it against the live product before any launch or campaign that points traffic at it.
• If you compare against competitors, compare named plans at a stated date, use each rival's published plan terms, and keep the basis of comparison visible. A table that implies a rival lacks a feature it has is a deceptive claim, not a copy win.

Full table craft, cell vocabulary, and a worked table: references/comparison-tables.md.

5. Use anchors and decoys with a clear limit

Anchoring and decoy effects are real, and they are easy to abuse. Use them to make a true value relationship easier to see. Never use them to invent value that is not there.

Anchoring:

• The first price a buyer sees sets the reference for the prices after it. Tversky and Kahneman described this as anchoring with insufficient adjustment in 1974: people start from an initial value and move too little from it.
• Honest anchors: a genuine higher plan with real added scope, a real former price that meets the test in the Operating Rules, the cost of the current workaround stated with its source, or the total cost of buying the parts separately when they are truly sold separately.
• Dishonest anchors: a struck-through price the product was never sold at, a premium plan built only to sit on the page, or a "value" total made by adding list prices nobody pays.
• Place the anchor next to the plan it is meant to frame, and label what it is. A buyer should be able to tell what the crossed or compared number refers to without a footnote hunt.

Decoys:

• Huber, Payne, and Puto described the asymmetric dominance effect in the Journal of Consumer Research in 1982: adding an option that is clearly worse than one existing option, but not clearly worse than the other, shifts choice toward the option that dominates it.
• The familiar illustration, popularized by Dan Ariely, is a subscription offered as digital alone, print alone at the same price as print plus digital, and the combined option. Print alone exists in that setup mainly to make the combined option look like obvious value.
• On a pricing page, apply the underlying principle only in its honest form: a real plan, priced for a real segment, can make the step up to the next plan legible by showing what a small price difference buys. The lower plan must still be a fair buy for the buyer it names.
• Never ship a plan you hope nobody buys, priced to lose on every dimension that matters, with the intent to push buyers up. That is a decoy in the abusive sense. It also creates support, refund, and review costs when someone does buy it.
• Test for abuse before publishing: if the plan were the only one a segment saw, would you defend its price as fair for what it delivers? If not, cut it or reprice it.

Middle-plan effect:

• Research on the compromise effect (Simonson, 1989) shows an option gains share when it sits in the middle of a set, because the middle is easy to justify. Expect the middle plan to draw the most buyers in a three-plan set. Price and scope it as the plan you actually want most buyers on, and make sure its limits fit the median customer, not a best-case customer.

6. Write FAQ that handles the real price objections

FAQ on a pricing page has one job: answer the objections that stop a purchase, in the order buyers raise them. Pull the questions from sales calls, support tickets, trial exit surveys, and search queries that land on the page. Do not invent questions nobody asks so the answers can sell.

Cover these when they apply, in this order:

1. Billing: when am I charged, what does the total come to, what happens at renewal, how do I get an invoice or receipt.
2. Switching: can I move up or down later, is the change prorated, do I lose work or data if I move down.
3. Cancellation: how do I cancel, what happens at the end of the paid period, is any part refundable.
4. Limits and overage: what counts toward the limit, what warning do I get, what does extra use cost, can I cap it.
5. Trials and guarantees: what the trial includes, whether a card is required, what happens when it ends, what the guarantee covers.
6. Discounts: who qualifies, how long a discount lasts, what the price becomes after it. State the full price after any discount period.
7. Security and procurement, for business plans: data handling, access controls, contract terms, and where the detailed documents live.
8. Fit: who a plan is not for. One honest not-for answer saves returns and builds trust with the buyers it is for.

Write each answer in 2 to 4 short sentences. Give the number, date, or condition in the answer itself. "Yes, you can cancel anytime" is incomplete without what happens to access and billing after cancellation. Keep answers consistent with the cards and the table: if the three places disagree, a buyer believes the worst one.

Question bank and answer patterns: references/objection-faq.md.

7. Audit the page before it ships

Run the audit tool on a plain-text or HTML draft, then do the human checks the tool cannot do.

• Run python3 scripts/pricing_page_audit.py <draft-file> from this skill folder. It flags missing billing periods, prices with no period nearby, "contact sales" with no starting price, blank comparison cells in markdown tables, struck-through prices with no former-price note, and discount claims with no post-discount price.
• Read every price aloud and reproduce it as a buyer: base, fees, billing period, renewal. If you cannot reach the same total the card shows, fix the card.
• Check each claim against its source: former price, competitor price, customer count, savings percentage. Remove any claim you cannot point to a record for.
• Read the page on a narrow screen. The recommended plan, the price, and the billing period must survive without the table.
• Test one variable at a time after launch: order, recommended plan, fit statement, or toggle default. Decide the metric and the run length before the test starts.

Output Contract

When asked to write or rewrite a pricing page, return:

1. The page copy in order: headline and subhead, plan cards, summary comparison, full comparison table, FAQ, final call to action.
2. For each plan: name, fit statement, full price with period and total, the deciding differences, and the call to action.
3. A claims list: every price claim, savings claim, former price, competitor claim, and proof claim on the page, each with its source or marked unverified and removed from the copy.
4. Open risks: missing facts, places where the cards, table, and FAQ could disagree, and the first test to run after launch. For a review of an existing page, return findings first, ordered by how much each one blocks a buyer from choosing, then the rewritten sections.

Operating Rules

1. No invented numbers. Prices, limits, savings, counts, and dates come from the product record or they do not appear.
2. A former price claim must pass the FTC test in the Guides Against Deceptive Pricing (16 CFR Part 233): the former price was real, openly offered, and held for a reasonably substantial period in the recent regular course of business, in good faith. If it was not, do not show it.
3. A competitor price comparison must be based on the rival's actual current price at which substantial sales are made, for a comparable plan, checked at a stated date. Otherwise cut the comparison.
4. Show the price a buyer can actually pay, including mandatory fees, with the billing period attached. Introductory prices always carry the renewal price.
5. Every plan on the page must be a fair buy for the segment it names. No plan exists only to make another plan look good.
6. One primary action per page. The recommended plan gets one clear mark, not a stack of badges.
7. Cards, table, and FAQ must agree. When they conflict, fix the source fact first, then all three places.
8. Mark assumption-based copy plainly when the underlying pricing facts were not supplied.

References

• Comparison tables: row selection, cell vocabulary, grouping, and a worked honest table.
• Objection FAQ: question bank by objection type, answer patterns, and answers to avoid.
• Worked pricing pages: two full pages, one self-serve and one with a sales-led tier, with notes on the choices made.
• Page anatomy: section order for self-serve, sales-led, and usage-based pages, plus toggle and annual billing copy.