ProductAdvanced8 min read

Pricing Experimentation

Pricing experimentation is the disciplined practice of testing prices, packaging, and value metrics with real prospects to find the price that maximizes long-term revenue, not just conversion. Pricing is the highest-leverage product decision: a 1% improvement in price typically lifts operating profit by 11% (vs. 3-7% for cost or volume improvements per McKinsey). Yet most companies test pricing exactly once — at launch — and never again. The companies that test continuously (Adobe, Netflix, Atlassian) extract 20-40% more LTV per customer than peers who set-and-forget.

Also known asPrice TestingPricing A/B TestsWillingness-to-Pay TestingVan WestendorpPricing Iteration

Challenge a friend Browse library

The Trap

The trap is conflating 'A/B price test' with 'change the price on the pricing page and watch conversion.' Real pricing experiments require: (1) controlling for cohort effects (prospects in different months convert differently), (2) measuring LTV not just conversion (cheaper prices convert better but produce worse customers), (3) protecting brand trust (existing customers seeing different prices than new ones triggers churn). Naïve A/B price tests produce data that misleads the company into lowering prices repeatedly, eroding margin. The other trap: copying competitor pricing as 'experimentation.' That isn't experimentation — it's anchoring on someone else's mistake.

What to Do

Run pricing experiments in this order: (1) Qualitative first — Van Westendorp surveys with 100+ ICP-matched prospects to find the acceptable price range. (2) Quantitative test new prospects only — never re-price existing customers in an experiment. (3) Test packaging, not just price level — moving a feature between tiers often beats changing the dollar amount. (4) Measure to 6-month LTV, not Day-1 conversion. (5) Run experiments quarterly, not annually. Most pricing pages stale-date within 12-18 months as the market and your value prop change.

Formula

Price Elasticity = (% Change in Quantity Sold) ÷ (% Change in Price). Inelastic (<1) means raise price.

In Practice

Adobe's 2013 shift from boxed Creative Suite ($1,800-2,600 one-time) to Creative Cloud subscription ($50/month, ~$600/year) is one of the most consequential pricing experiments in SaaS history. The transition was tested in stages — early Creative Cloud existed alongside boxed Suite for 18 months while Adobe measured cohort behavior, churn, and total spend. The data showed subscription customers spent more over 3 years than boxed customers AND gave Adobe predictable recurring revenue. The fully subscription move launched in 2013; Adobe stock 5x'd over the following 5 years and revenue more than tripled. (Source: Adobe annual reports, 2012-2018)

Pro Tips

01
Test packaging changes before testing price changes. Moving a high-value feature from Pro to Business often produces 15-25% price-point lift with less risk than a price increase.
02
Grandfather existing customers when raising prices. The PR cost of re-pricing existing accounts is far higher than the revenue gained — and they'll churn at higher rates anyway.
03
Test pricing on the 'invisible' parts of the page first — annual vs. monthly toggle, default selection, tier ordering. These changes often produce 5-15% revenue lift with no actual price movement.

Myth vs Reality

Myth

“You can A/B test prices on your pricing page like any other element”

Reality

Public price A/B tests violate trust if discovered (people compare on Reddit, screenshots leak). Most companies test prices via segmented landing pages, geographic experiments, or sales-quoted prices for B2B — not the public pricing page itself. Stripe and others have explicit guidance against displayed-price A/B testing.

Myth

“Higher conversion at a lower price means you should lower the price”

Reality

Lower-price cohorts often have worse retention, lower expansion, and higher support costs. Day-1 conversion is the wrong metric. Price experiments must measure 6-12 month LTV to avoid optimizing toward 'cheap users who don't stay.'

Try it

Run the numbers.

Pressure-test the concept against your own knowledge — answer the challenge or try the live scenario.

🧪

Scenario Challenge

You ran a 30-day price test: half of new visitors saw $99/month, half saw $149/month. The $99 cohort converted 22% more visitors. The $149 cohort generated 18% more revenue per visitor. Marketing wants to ship the $99 price for higher conversion. What's the right call?

Industry benchmarks

Is your number good?

Calibrate against real-world tiers. Use these ranges as targets — not absolutes.

Pricing Review Cadence (mature SaaS)

Mature B2B SaaS pricing reviews

Best Practice

Every 6-12 months

Acceptable

Every 12-18 months

Stale

Every 2-3 years

Set-and-Forget

3+ years unchanged

Source: Hypothetical: industry observation aggregated across pricing consultancies (Simon-Kucher, OpenView)

Real-world cases

Companies that lived this.

Verified narratives with the numbers that prove (or break) the concept.

🎨

Adobe

2011-2013

success

Adobe ran a multi-year pricing experiment moving from $1,800-2,600 boxed Creative Suite to $50/month Creative Cloud subscription. The transition was staged: Creative Cloud launched in 2011 alongside boxed Suite. For 18 months Adobe measured how each cohort behaved — retention, total spend, feature usage, support load. Subscription customers spent more over a 3-year window AND produced predictable recurring revenue. In 2013 Adobe killed boxed Creative Suite entirely. Stock price 5x'd over the next 5 years; revenue went from $4.4B (2012) to $15.8B (2020).

Pre-Subscription Revenue (2012)

$4.4B

Post-Subscription Revenue (2020)

$15.8B

Stock Performance (2013-2018)

~5x

Test Window

18 months parallel pricing

Major pricing changes deserve major experiments. Adobe didn't just flip the switch — they ran the experiment for 18 months and let cohort data make the call. Discipline at the pricing-strategy level produces decade-defining outcomes.

Source ↗

Decision scenario

The Price Increase Decision

You're VP Product at a $20M ARR SaaS company. Your prices haven't changed in 3 years. Customer NPS is 45, churn is 1.5%/month. Internal analysis shows your competitors have raised prices 20-30% over the same period. The CFO wants a 25% price increase across all customers. Sales is terrified.

ARR

$20M

Avg Customer Price

$200/month (3 years unchanged)

Monthly Churn

1.5%

NPS

45 (healthy)

Competitor Pricing

20-30% higher

Decision 1

The CFO's plan: raise all customers to $250/month at next renewal. Sales argues this will trigger a churn wave. You have to recommend an approach to the CEO.

Endorse the CFO's plan — raise all customers 25% at renewal, accept the churn riskReveal

At renewal, churn doubles from 1.5% to 3.2% for the first 3 months. Roughly 12% of customers leave during the renewal cycle. The remaining 88% pay 25% more, so revenue lifts ~10% net. But CSAT drops, support volume spikes (negotiation requests), and new sales get harder because departed customers warn the market. Net financial outcome: positive but smaller than projected, and the brand cost lingers.

Customer Count: −12%ARR: +10%Brand: Damaged

Grandfather existing customers at current pricing for 12 months; raise prices 25% only for new customers immediately; revisit existing-customer pricing after measuring new-customer cohort behaviorReveal

New customers convert at the higher price with only a small dip (5% conversion drop vs. 25% price lift = highly profitable). Existing customers feel respected and don't churn. Six months in, you have data showing the new price works. You then introduce a smaller existing-customer increase (10-12%) tied to product improvements shipped during the year. Net: 18-20% ARR lift over 12 months with minimal churn. Brand strengthens.

Customer Count: StableARR: +18-20% over 12moBrand: Strengthened

Related concepts