Marketing Skill

Pricing Strategy

The pricing skill helps an AI agent decide what to charge and how to package it: value metric, pricing model, tiers, freemium versus free trial, price increases, and pricing page teardowns. Use it when you're setting a first price, your pricing feels arbitrary, or you want to know whether AI assistants can read your pricing page.

Maintained by Corey Hainesv2.1.7Updated 2026-11-094 references
npx skills add coreyhaines31/marketingskills

Installs the full library — Pricing Strategy included

SaaS pricing comes down to three decisions: packaging (what each tier includes), the value metric (what you charge for), and the price point. Price on value, not cost or a competitor's number: land between the next-best alternative and the customer's perceived value, aiming for roughly 10x perceived value per dollar paid. Choose a metric that grows as the customer gets more value, then package it good-better-best, with the middle tier as the anchor and the top tier around 2–3x it. On day one, pick a price you can learn from: about $10/mo for individuals, $100/mo for SMB teams, $1,000/mo for mid-market, and avoid the $9 trap of false traction. Test increases on new customers first, then move 5–10% of existing accounts before expanding.

How the Pricing Strategy skill works

  1. 1

    Gather business and performance context

    The agent reads your product marketing context, then fills gaps: product type, current pricing, SMB, mid-market, or enterprise, and self-serve, sales-led, or hybrid motion. It asks what value you deliver, which alternatives buyers weigh, how competitors price, and your current conversion rate, ARPU, and churn. It also asks whether you're optimizing for growth, revenue, or profitability, since that changes every later choice.

  2. 2

    Find the value metric

    One test decides the metric: as a customer uses more of it, do they get more value? Seats suit collaboration tools, usage suits variable consumption, contacts suit CRM and email tools, and flat fees suit simple products. The metric then points to one of eight models: flat, usage, tier, user, feature, credit, outcome-based, or a hybrid, which is where most mature products end up.

  3. 3

    Set or research the price point

    With no data, the agent starts from the $10, $100, or $1,000 per month bucket that matches the customer and steers away from ultra-low prices. With customers to survey, it runs Van Westendorp with 100–300 respondents segmented by persona, MaxDiff to rank features, or Gabor-Granger and conjoint for demand. Survey intersections give candidate prices to test, not proof of revenue impact.

  4. 4

    Package good-better-best tiers

    Tiers differ by feature gating, usage limits, support level, or access such as API, SSO, and custom branding. Three tiers is the norm; five or more risks decision paralysis. The agent weighs freemium against a free trial (7–14 days for simple products, 14–30 for complex ones) and adds a Contact Sales tier only when deals exceed about $10k ARR or need custom contracts.

  5. 5

    Plan price increases as rollouts

    Signals that it's time include conversion above 40%, monthly churn under 3%, and “it's so cheap” feedback. The rollout raises prices for new signups first, uses grandfathering as a grace period instead of a permanent exemption, moves 5–10% of existing customers before staggered waves, and explains the why months ahead with a generous option such as locking the old price with an annual upgrade.

  6. 6

    Tear down the pricing page

    For an existing page, the agent scores 10 dimensions as Pass, Partial, or Gap. Six cover the human buyer, including value-prop clarity, cognitive load, and transparency. Four cover AI-agent readiness: prices in text, extractable FAQs, per-tier limits stated in words, and structured data. It runs a paste test with a browsing AI, then ranks fixes by impact and effort.

When to use it

Your agent reaches for this skill when you ask for things like:

“pricing”“pricing tiers”“freemium”“free trial”“packaging”“price increase”“value metric”“Van Westendorp”“willingness to pay”“monetization”

What's inside

The playbook covers

Before StartingPricing FundamentalsInitial Pricing — "Pick a Price You Can Learn From"Value MetricsTier Structure OverviewPricing ResearchWhen to Raise PricesPricing Page Best PracticesPricing Page TeardownPricing ChecklistTask-Specific Questions

See it in action

You ask

Help me figure out pricing for our new SaaS product. It's a customer support platform for e-commerce stores. We're not sure whether to charge per agent, per ticket, or flat rate. Currently thinking $49-199/month range.

The skill delivers
Value metric check

Ask of each option: as a store uses more of it, does it get more value? Per agent: weak. A 2-person store can handle thousands of tickets, and seats punish adoption (teams share logins to avoid paying). Per ticket: tracks value, since more orders mean more tickets, but buyers fear a surprise bill during holiday spikes. Needs caps or alerts. Flat rate: easiest to sell and forecast, but undercharges your largest stores. Working pick: tiers built on a monthly ticket allowance (usage limits), with seats as a secondary limit. If you later add AI auto-resolution, an outcome-based fee per resolved ticket can sit on top as a hybrid.

Draft tiers inside your $49–199 range

Starter: $49/mo, up to 1,000 tickets/mo, 2 agents, email and chat inbox, store integration Growth: $99/mo, recommended anchor, up to 3,000 tickets/mo, 5 agents, macros, order data in the ticket view Scale: $199/mo, up to 8,000 tickets/mo, unlimited agents, SLAs, API access, priority support Alerts at 80% of the allowance, no automatic overage charges. Annual billing at about 17% off. This sits in the ~$100/mo SMB bucket, safely above the $9 trap.

Validate before launch

Start with real buyer conversations about what stores pay for support tools today. Then run Van Westendorp with 100–300 store owners, segmented by monthly order volume. Treat the intersections as candidate prices, then test the Growth price on new signups only and watch conversion. Check value/price: Growth should be worth roughly $1,000/mo to a store, about 10x its price. If you can't articulate that, fix the offer or positioning before the price.

More things to ask

Real requests the skill is tested against:

  • We want to raise our prices by 30%. We've been at $29/month for 2 years and we've added a lot of features. How do we do this without losing customers?

  • how do we figure out what people will actually pay? we're launching a new product and have no idea what to charge.

  • We have a Basic ($19), Pro ($49), and Enterprise (custom) plan. The Pro plan gets 70% of signups. Should we add a plan between Pro and Enterprise?

  • What pricing psychology tactics should we use on our pricing page? We want the $79 plan to be the most popular.

  • Our pricing page conversion rate is only 1.5%. Can you review the page and suggest improvements?

FAQ

How much should I charge for a new SaaS product?

On day one, pick a price you can learn from rather than a precise one. Start near the order of magnitude that matches your buyer: about $10/mo for prosumers, $100/mo for SMB teams, and $1,000/mo for mid-market or business-critical tools. Resist ultra-low prices like $9/mo. They create false traction and are hard to raise 5–10x later. Round and slightly higher beats clever and cheap.

Should I charge per seat, per usage, or a flat rate?

Let the value metric decide. Per seat fits only when value rises with headcount; otherwise teams share logins. Usage pricing is the fairest when consumption tracks value, but revenue is less predictable and buyers fear surprise bills, so pair it with caps or alerts. Flat rate is the easiest to sell and forecast but leaves money on the table with large customers. Many products end up hybrid: a base fee plus usage.

How does the Van Westendorp price survey work?

You ask 100–300 target buyers four questions: at what price is it too expensive, so cheap you'd question quality, getting expensive, and a bargain. Plot the four curves, cheap curves falling and expensive curves rising, and read the intersections. The acceptable range runs from marginal cheapness to marginal expensiveness. The “optimal price point” is only a survey intersection; it doesn't measure conversion, churn, or demand, so validate with real price tests.

Should I offer freemium or a free trial?

Freemium fits products with network effects, free users who add value, low cost to serve, and a clear limit that triggers upgrades. A free trial fits products that need setup or time to show value, sell to buying committees, or carry higher prices. Use 7–14 days for simple products and 14–30 for complex ones. Requiring a card upfront lifts trial-to-paid conversion (roughly 40–50% vs. 15–25%) but lowers trial volume.

How do I raise prices without losing customers?

Treat it as a rollout. Raise the price for new customers first, since they have no anchor. Don't grandfather forever: a $50/mo customer who should pay $250/mo is a $2,400/yr gap. Move 5–10% of existing accounts, watch churn and support volume for a cycle, then expand in waves. Explain the reason months ahead and offer a generous option. Expect some churn from your most price-sensitive accounts.

Can AI assistants read my pricing page?

Often not. Prices in an image, rendered only by JavaScript, behind a login, or replaced by “Contact us” give text-fetching agents nothing to quote, so they quote a competitor instead. Run the paste test: ask a browsing AI for your plans and prices and see what it misses. Fixes include prices in plain text, limits stated in words, an FAQ, Product and Offer schema, and letting AI search bots crawl.

Related skills

Use it in your agent

Pricing Strategy is a plain SKILL.md folder, so the same skill loads in any agent that reads the skills format. Setup guides:

Build a tool for this workflow?

Marketers run this skill mid-task — the exact moment your tool matters. Partner tiers put your tool inside the skill their agent runs.

See partner tiers