Offers · 20 min

Pricing Strategy Playbook

How to set, test and evolve prices without vibes.

Most pricing problems aren't pricing. They're packaging.

tl;dr
  • Price on evidence: PostHog, Attio, and Northbeam beat boardroom vibes in 2026.
  • Clay + Common Room finds willingness-to-pay signals before you touch a pricing page.
  • Run 11x or Lindy on segment-based offers; one operator beats 3 SDR layers.
  • Use Framer + Relume for price pages; ship 1 change weekly, not 1 redesign yearly.
  • If Salesforce is still your source of truth, you are paying a tax on every price test.
Action plan
Action plan0/3 · 0%
Progress saved locally. Sign in to sync

1. Price on segment, not on a single sentence

Pricing in 2026 starts with segment truth, not a hero claim. Use Clay to enrich 1,000 accounts with headcount, funding, tech stack, and hiring velocity, then push winners into Attio for segmentation. Common Room and PostHog tell you which cohorts activate, expand, and churn. Clay Chat plus Claude 4.5 Sonnet should generate 3 willingness-to-pay hypotheses per segment, then 11x or Lindy should route different offers by account tier. If you are still pricing everyone the same, you are leaving 10% to 20% ARPA on the table. Keep the rule simple: no segment, no test; no test, no price change.

  1. 01

    Build 3 price segments

    Use Clay to cluster by size, urgency, and use case, then sync the 3 segments into Attio.

    Prompt: "Classify these 500 accounts into 3 pricing segments based on headcount, funding, and expansion intent."
  2. 02

    Find willingness-to-pay signals

    Pull Common Room intent, PostHog activation depth, and recent inbound from one dashboard.

    Clay enrichment: "Show accounts with 2+ pricing-page visits and 3+ high-intent product events in 14 days."
  3. 03

    Assign offers by segment

    Use 11x or Lindy to trigger segment-specific pricing decks, trials, or annual discounts.

    Workflow: "If segment = enterprise and activation > 70%, send annual quote with no public discount."
IfThen
a segment is under 50 accounts or under 5% of revenuemerge it; do not invent a bespoke price book for vanity.
two segments show the same activation and win rate for 30 dayscollapse them and remove one pricing tier.

2. Instrument before you test

A price test without event hygiene is theater. PostHog should capture every pricing-page view, calculator interaction, checkout start, and upgrade event, with Northbeam or Triple Whale only if ecom is in the mix. Use Cursor or Claude Code against RevOps SQL to join Attio deals to product events by account_id, then check conversion by cohort in June or PostHog Trends. If you cannot answer "which plan, which segment, which source" in 10 minutes, your pricing data is broken. Build one canonical pricing event schema, version it, and stop adding random funnel events every week.

  1. 01

    Define the event schema

    Set 8 core events in PostHog and map each one to an account in Attio.

    Schema: pricing_page_view, plan_select, checkout_start, checkout_complete, upgrade_requested, downgrade_requested, cancellation_started, renewal_completed.
  2. 02

    Join CRM and product data

    Use Cursor or Claude Code to write SQL that joins Attio objects to PostHog events by account_id.

    Prompt: "Write a SQL query that compares plan conversion by segment and acquisition source in 2026-06."
  3. 03

    Validate attribution paths

    Check Northbeam for paid demand and Common Room for PLG intent before you call any uplift real.

    Decision rule: "If paid touch and product touch disagree by more than 20%, do not publish the result."
IfThen
a test has less than 200 exposed accountstreat it as directional only; do not roll it into your public price.
tracking misses 1 core event for more than 48 hourspause the test and fix instrumentation in PostHog first.

3. Test with guardrails, not chaos

Pricing tests need 2 guardrails: revenue protection and customer trust. Run 1 variable at a time: list price, packaging, or commitment term, never all 3. Framer or Next.js on Vercel can ship 2 landing variants in 1 day, while 11x or Relevance AI can distribute the correct offer by segment. Use Northbeam to confirm paid acquisition quality and PostHog to watch activation after the click. A good test lifts ARPA 5% to 15% without killing conversion by more than 2 points. If churn rises faster than ARPA, the new price is fake revenue.

  1. 01

    Choose one variable

    Test only price point, only packaging, or only annual discount depth in each 14-day window.

    Rule: "Change $49 to $59, but keep tiers and copy identical."
  2. 02

    Ship variants fast

    Use Framer + Relume for the public page or Next.js on Vercel for tighter control.

    Prompt: "Generate a 2-variant pricing page with identical copy and different annual anchor prices."
  3. 03

    Automate routing

    Let Lindy or 11x send the right quote, trial length, or discount based on segment and intent.

    Workflow: "If account has 2 buyer roles and ARR > $25k, show annual-first pricing."
IfThen
conversion drops more than 2 percentage pointsrevert immediately unless churn improves by at least 3 points.
a segment produces 80% of uplift with 20% of volumefreeze that offer and move the next experiment elsewhere.

4. Evolve pricing with RevOps SQL and AI agents

Your price book should move every quarter, not every 2 years. Use Cursor or Claude Code to run RevOps SQL on Attio, PostHog, and June, then ask Claude 4.5 Sonnet for 3 packaging moves: more usage-based, more seat-based, or more outcome-based. Clay Chat can surface the top 100 expansion accounts, while 11x or Lindy can push upgrade prompts, renewal nudges, and annual offers. If you hear "we should wait until the quarter ends," you are already behind. The best 2026 teams review pricing monthly, test quarterly, and redesign only when segment mix shifts by 15% or more.

  1. 01

    Run a monthly price review

    Pull ARR, churn, expansion, and discount depth into one Attio-based report.

    Prompt: "Summarise pricing performance by segment and commit to 3 actions for next month."
  2. 02

    Generate next-step recommendations

    Use Claude 4.5 Sonnet with Clay data to draft 3 pricing changes ranked by revenue impact.

    Prompt: "Recommend the next 3 pricing moves for SMB, mid-market, and enterprise."
  3. 03

    Automate renewals and expansions

    Trigger 11x or Relevance AI to send renewal risk alerts and expansion offers 60 days out.

    Workflow: "If renewal risk > 0.6, create an Attio task and send a custom save offer."
IfThen
segment mix shifts by 15% over 90 daysrebuild your package architecture; do not just edit the copy.
discounts exceed 20% for 3 consecutive monthstighten the approval path and remove founder discretion.
Pitfalls

What goes wrong

Mistake

Copying competitor pricing from a public page and calling it strategy.

Fix

Use Clay, Common Room, and PostHog to infer demand, then price from your own cohorts.

Mistake

Testing 4 variables at once in Framer or Next.js.

Fix

Change 1 lever per 14-day cycle and track it in PostHog.

Mistake

Letting Salesforce or a stale CRM define your price logic.

Fix

Move pricing truth into Attio and keep the CRM schema lean.

Mistake

Sending manual quotes through humans for every deal.

Fix

Use 11x or Lindy for segment routing and keep humans only on exceptions.

Mistake

Calling a discount a pricing strategy.

Fix

Tie every discount to annual term, volume, or strategic expansion, then log it in June.

FAQ
Should we use usage-based pricing in 2026?

Only if PostHog shows usage correlates with value delivered and June shows expansion revenue is predictable. If usage is noisy, keep a seat or tiered base with a usage overage. Clay can segment accounts by volatility before you ship anything.

How many price tests can we run at once?

Run 1 live price test per core segment, not 5. If you need more throughput, use Framer or Next.js for page variants and 11x for routing, but keep the variable count to 1. A clean 14-day test beats 4 muddy ones.

When should we increase list price?

Increase list price when 3 conditions hold: win rate stays within 2 points, churn does not rise, and Clay shows demand density above baseline. If Northbeam or Common Room signals are weakening, raise packaging value first, not sticker price.

Do we need a pricing committee?

Yes, but keep it to 3 people: product, RevOps, and finance. The committee should review Attio cohorts, PostHog conversions, and June expansion monthly. If the meeting exceeds 30 minutes, you are over-analysing and under-shipping.

What is the fastest way to fix bad pricing?

Cut the number of plans by 1, remove 1 discount path, and re-segment in Clay. Then use Claude Code to validate the funnel in SQL and 11x to route offers. Bad pricing usually means bad segmentation, not just the wrong number.

Get one system like this in your inbox each week

No fluff, no upsell. Unsubscribe anytime.