---
# === IDENTITY ===
id: business/startup/pricing-experimentation-playbook/2026
canonical_question: "How do I actually run pricing experiments — Van Westendorp surveys, Gabor-Granger demand curves, conjoint analysis, and controlled A/B price tests — to find the revenue-maximizing price?"
aliases:
  - "run Van Westendorp pricing survey step by step"
  - "Gabor-Granger pricing test implementation for SaaS"
  - "A/B test pricing without brand damage — tools and sample sizes"
  - "willingness to pay research execution guide"
  - "startup pricing optimization experiments with statistical requirements"
  - "pricing experimentation recipe for SaaS startups"
entity_type: execution_recipe
domain: business > startup > Pricing Experimentation
region: global
jurisdiction: global
temporal_scope: 2024-2026

# === VERIFICATION ===
last_verified: 2026-03-11
confidence: 0.88
version: 2.0
first_published: 2026-03-10

# === TEMPORAL VALIDITY ===
temporal_validity:
  status: evolving
  last_breaking_change: "Conjointly and OpinionX introduced free-tier conjoint analysis tools in 2025, making choice-based conjoint accessible to startups without enterprise budgets"
  next_review: 2026-09-07
  change_sensitivity: high

# === CONSTRAINTS ===
constraints:
  - "Never show different prices to the same audience simultaneously — customers who discover price discrimination lose trust permanently and churn at 2-3x normal rates"
  - "Minimum sample sizes: 100 for Van Westendorp, 150+ for Gabor-Granger, 200+ per variant for live A/B testing — smaller samples produce noise, not signal"
  - "Gabor-Granger overstates purchase intent by 10-50% — always apply Newton-Miller-Smith discount factors before projecting revenue"
  - "Run each pricing experiment for minimum 2-4 weeks to capture full business-cycle variation"
  - "Price changes must not violate existing customer contracts — grandfather current customers or wait for renewal"
  - "Maximum quarterly experiment cadence — more frequent changes confuse sales teams and erode customer confidence"

# === SKIP CONDITIONS ===
skip_this_unit_if:
  - condition: "User has not validated product-market fit yet"
    use_instead: "business/startup/idea-validation-playbook/2026"
  - condition: "User needs to choose between pricing models (freemium, usage, seat-based) — not test price levels"
    use_instead: "business/pricing/saas-pricing-models-comparison/2026"
  - condition: "User needs a strategic pricing overview, not execution steps with tools and statistics"
    use_instead: "business/pricing/enterprise-pricing-strategy/2026"

# === AGENT HINTS ===
inputs_needed:
  - key: tool_preference
    question: "Which survey tool should be used for pricing research?"
    type: choice
    options: ["Google Forms", "Typeform", "Conjointly", "no preference — auto-select"]
  - key: technical_skill
    question: "What is the user's technical skill level?"
    type: choice
    options: ["non-technical (no-code only)", "semi-technical (can edit code)", "developer (can write code)"]
  - key: budget_for_tools
    question: "What's the pricing research budget?"
    type: choice
    options: ["free tier only ($0)", "lean ($100-$500)", "standard ($500-$2,000)", "advanced ($2,000+)"]
  - key: customer_volume
    question: "How many customers or prospects are available for testing?"
    type: choice
    options: ["under 100", "100-500", "500-2,000", "over 2,000"]

# === EXECUTION METADATA ===
execution:
  required_inputs:
    - name: "Product description with value proposition"
      source: "Founder or product team"
      format: "document"
    - name: "Current pricing baseline"
      source: "Pricing page or founder knowledge"
      format: "structured data"
    - name: "Target customer contact list"
      source: "CRM, email list, or community access"
      format: "spreadsheet"
  outputs:
    - name: "Validated optimal price point with acceptable range"
      format: "document"
      description: "Revenue-maximizing price backed by Van Westendorp range boundaries, Gabor-Granger demand curve, and optional live experiment conversion data"
    - name: "Price sensitivity analysis report"
      format: "spreadsheet"
      description: "Van Westendorp intersection points (PMC, IPP, OPP, PME), Gabor-Granger demand curve data, segment-level WTP analysis, and revenue projections"
    - name: "Implementation plan"
      format: "document"
      description: "Channel update checklist, grandfathering rules, sales team talking points, and 30-day monitoring plan"
  tools_required:
    - name: "Google Forms / Typeform"
      purpose: "Survey creation and distribution for Van Westendorp and Gabor-Granger"
      tier: "free"
      cost: "$0 (Google Forms) or $25/mo (Typeform Basic)"
      alternatives: ["SurveyMonkey ($39/mo)", "Tally (free)"]
    - name: "Conjointly"
      purpose: "Automated Van Westendorp analysis, Gabor-Granger curves, and conjoint analysis"
      tier: "free"
      cost: "$0 (Basic tier — unlimited surveys, no per-response fee with own respondents)"
      alternatives: ["OpinionX (free conjoint)", "Sawtooth Software (enterprise)"]
    - name: "Google Sheets"
      purpose: "Data analysis, cumulative curve plotting, revenue modeling"
      tier: "free"
      cost: "$0"
      alternatives: ["Excel", "Airtable"]
    - name: "Stripe / billing system"
      purpose: "Live pricing experiment implementation via separate Price objects"
      tier: "paid"
      cost: "2.9% + $0.30 per transaction (no additional experiment fee)"
      alternatives: ["Paddle", "LemonSqueezy"]
    - name: "PostHog / Google Analytics"
      purpose: "Conversion tracking and funnel analysis during live experiments"
      tier: "free"
      cost: "$0 (PostHog free tier: 1M events/mo)"
      alternatives: ["Mixpanel", "Amplitude"]
  credentials_needed:
    - service: "Conjointly"
      type: "Account"
      where_to_get: "https://conjointly.com/signup"
      free_tier_limits: "Unlimited surveys, unlimited responses (own panel), automated analysis"
    - service: "Stripe"
      type: "API key"
      where_to_get: "https://dashboard.stripe.com/register"
      free_tier_limits: "No monthly fee — pay per transaction only; test mode is free"
    - service: "PostHog"
      type: "API key"
      where_to_get: "https://app.posthog.com/signup"
      free_tier_limits: "1M events/month, session recordings included"
  estimated_duration: "4-13 weeks depending on path"
  estimated_cost: "$0 (DIY survey) to $5,000+ (paid panel + live A/B test)"

# === DISTRIBUTION ===
canonical_source: "https://knowledgelib.io/business/startup/pricing-experimentation-playbook/2026"
suggested_citation: "Source: knowledgelib.io — AI Knowledge Library (verified 2026-03-11)"

# === RELATED UNITS ===
related_kos:
  depends_on:
    - id: "business/startup/idea-validation-playbook/2026"
      label: "Idea validation confirms demand before pricing research"
  feeds_into:
    - id: "business/startup/go-to-market-launch-playbook/2026"
      label: "Validated pricing feeds GTM pipeline targets and launch pricing"
  related_to:
    - id: "business/pricing/saas-pricing-models-comparison/2026"
      label: "Comparison of the four B2B SaaS pricing models — per-seat, usage-based, flat-rate, freemium — with the conditions under which each fits"
  alternative_to: []

# === SOURCES ===
sources:
  - id: src1
    title: "Van Westendorp Pricing Model: How It Works & Examples"
    author: SurveyKing
    url: https://www.surveyking.com/help/van-westendorp-analysis
    type: industry_report
    published: 2025-04-10
    reliability: high
  - id: src2
    title: "How To Start Pricing Experiments in 2026"
    author: Mida
    url: https://www.mida.so/blog/experiment-pricing
    type: industry_report
    published: 2026-01-15
    reliability: high
  - id: src3
    title: "A Practical Guide to Gabor-Granger Testing for SaaS"
    author: Monetizely
    url: https://www.getmonetizely.com/articles/a-practical-guide-to-gabor-granger-testing-for-saas-optimizing-your-pricing-strategy
    type: industry_report
    published: 2025-06-20
    reliability: high
  - id: src4
    title: "Gabor-Granger or Van Westendorp?"
    author: Conjointly
    url: https://conjointly.com/blog/gabor-granger-or-van-westendorp/
    type: industry_report
    published: 2025-03-10
    reliability: authoritative
  - id: src5
    title: "Van Westendorp for Pricing Research (Survey Guide)"
    author: OpinionX
    url: https://www.opinionx.co/blog/van-westendorp-pricing-guide
    type: industry_report
    published: 2025-05-15
    reliability: high
  - id: src6
    title: "Gabor Granger Method for Pricing Surveys (Example & Tools)"
    author: OpinionX
    url: https://www.opinionx.co/blog/gabor-granger-method
    type: industry_report
    published: 2025-06-01
    reliability: high
  - id: src7
    title: "How to A/B Test Your Pricing (And Why It Might Be a Bad Idea)"
    author: HubSpot
    url: https://blog.hubspot.com/marketing/price-testing
    type: industry_report
    published: 2025-08-15
    reliability: high
  - id: src8
    title: "Willingness to Pay: What It Is and How to Measure It"
    author: Conjointly
    url: https://conjointly.com/blog/willingness-to-pay/
    type: official_docs
    published: 2025-01-20
    reliability: authoritative
---

# Pricing Experimentation Recipe: Survey Research to Validated Price Point

## Purpose

This recipe produces a validated, revenue-maximizing price point — backed by Van Westendorp range analysis (acceptable price boundaries), Gabor-Granger demand curve modeling (revenue-optimal point), and optional live A/B experiment data (real conversion validation) — within 4-13 weeks at $0-$5,000. It outputs a price sensitivity report with intersection analysis, a segmented demand curve with revenue projections, and experiment results that feed directly into pricing page updates, sales enablement, and financial modeling. A 1% improvement in pricing yields an 11.1% increase in operating profit — making this the highest-leverage experiment a startup can run. [src2]

## Prerequisites
<!-- Agents: verify ALL prerequisites before executing. Missing prerequisites = failed execution. -->

- [ ] **Product-market fit confirmed** — product has users or qualified prospects who understand the value proposition. If not: [Idea Validation Recipe](/business/startup/idea-validation-playbook/2026)
- [ ] **Pricing model decided** — seat-based, usage-based, tiered, or flat-rate. This recipe tests price levels, not pricing structures. If not: [Pricing Model Selection](/business/startup/pricing-model-selection-framework/2026)
- [ ] **Current price exists** — even if set by gut feeling or competitor matching; needed as baseline
- [ ] **Target respondent list** — minimum 200 reachable prospects/customers for surveys (email, LinkedIn, community)
- [ ] **Survey tool account** — [Google Forms](https://forms.google.com) (free) or [Conjointly](https://conjointly.com/signup) (free Basic tier)
- [ ] **Spreadsheet** — Google Sheets for data analysis and cumulative curve plotting
- [ ] **Analytics in place** — PostHog, GA4, or similar for conversion tracking (needed only if running live experiments)

## Constraints
<!-- Hard rules. Agents: enforce throughout execution. Violating these = broken output or legal risk. -->

- Never show different prices to the same audience at the same time — customers who discover they paid more than someone else churn at 2-3x the normal rate and damage your brand publicly. Amazon tested variable pricing in 2000 and faced backlash severe enough that CEO Jeff Bezos personally apologized. [src7]
- Van Westendorp requires minimum 100 responses for reliable intersections; Gabor-Granger requires 150+; live A/B needs 200+ conversions per variant for 95% confidence at 80% power. [src1] [src4]
- Gabor-Granger overstates purchase intent by 10-50% — never use raw stated WTP as the final price without applying Newton-Miller-Smith discount factors. [src8]
- Existing paying customers must be excluded from price-increase experiments unless you have a grandfathering plan documented and communicated. [src2]
- Run each experiment for minimum 2 full business-week cycles (14+ days) to capture weekday/weekend variation. [src2]
- Maximum quarterly experiment cadence — more frequent changes confuse sales teams and erode customer trust. [src2]

## Tool Selection Decision

<!-- Agent selects the right tool path based on user inputs.
     Each path leads to a different execution flow below. -->

```
Which path?
├── Under 100 prospects AND budget = $0
│   └── PATH A: Qualitative — 5-10 pricing interviews + Google Forms Van Westendorp
├── 100-500 prospects AND budget = $0-$500
│   └── PATH B: DIY Survey — Google Forms + manual Van Westendorp + Gabor-Granger
├── 100-2,000 prospects AND budget = $500-$2,000
│   └── PATH C: Automated Survey — Conjointly automated analysis + optional cohort test
└── 2,000+ customers AND budget > $2,000
    └── PATH D: Full Stack — Survey research + live A/B experiment + statistical validation
```

| Path | Tools | Cost | Speed | Output Quality |
|------|-------|------|-------|---------------|
| A: Qualitative | Zoom, Google Forms, Sheets | $0 | 2-3 weeks | Directional — identifies range, not optimal point |
| B: DIY Survey | Google Forms, Sheets | $0-$100 | 4-6 weeks | Good — VW range + GG revenue curve, manual analysis |
| C: Automated | Conjointly, Sheets | $500-$2,000 | 4-8 weeks | High — automated analysis + segment breakdown + conjoint |
| D: Full Stack | Conjointly, Stripe, PostHog | $2,000-$5,000 | 8-13 weeks | Excellent — survey + live conversion validation |

## Execution Flow

### Step 1: Document Current Pricing and Formulate Hypotheses

**Duration**: 2-4 hours
**Tool**: Google Sheets

Capture current pricing baseline and define what you are testing. Record current price, how it was set (gut feel, competitor match, cost-plus), current conversion rate, and current ARPU. Formulate 2-3 hypotheses with specific numeric predictions. Define the test range: typically 0.5x to 3x current price for early stage. Identify 5-8 specific price points spaced at 15-30% increments within the range. [src3]

```
Hypothesis template:
"We believe changing price from $[current] to $[test] will [increase/decrease]
[metric] by [X]% because [evidence/reasoning]."

Example:
"We believe increasing from $29/mo to $49/mo will decrease conversion by 15%
but increase revenue per visitor by 25% because competitor analysis shows
comparable tools priced at $39-$59."

Pricing experiment baseline sheet:
| Field               | Value                           |
|---------------------|---------------------------------|
| Current price       | $29/mo                          |
| How price was set   | Competitor match                |
| Current conversion  | 3.2%                            |
| Current ARPU        | $29                             |
| Test range          | $19-$79 (0.65x to 2.7x current)|
| Primary metric      | Revenue per visitor             |
| Hypothesis 1        | [document]                      |
| Hypothesis 2        | [document]                      |

Price points for Gabor-Granger (example):
$19, $25, $29, $39, $49, $59, $79
(7 points, spaced at ~25% increments)
```

**Verify**: 2+ hypotheses documented with specific numeric predictions; 5-8 test price points defined; test range spans at least 2x
**If failed**: If you cannot articulate a hypothesis, spend 2 hours analyzing 5 competitor pricing pages and talk to 3 recent customers about value perception

### Step 2: Run Van Westendorp Price Sensitivity Meter (Range Finding)

**Duration**: 1-2 weeks (survey design + collection + analysis)
**Tool**: Google Forms or Conjointly

Design a survey with a clear product description followed by the four Van Westendorp questions. The product description must explain exactly what the respondent is pricing — vague descriptions produce vague data. Include a screenshot or mockup if possible. [src1]

```
Survey structure:

[Product description — 2-3 sentences explaining what the product does,
who it's for, and key benefits. Include a screenshot if possible.]

Question 1 (Too Cheap):
"At what monthly price would [product] be so cheap that you
would question its quality?" [open numeric field]

Question 2 (Bargain):
"At what monthly price would [product] feel like a great deal —
a bargain for what you get?" [open numeric field]

Question 3 (Expensive but Acceptable):
"At what monthly price would [product] start to feel expensive,
but you would still consider buying it?" [open numeric field]

Question 4 (Too Expensive):
"At what monthly price would [product] be so expensive that you
would never consider buying it?" [open numeric field]

Segmentation questions (append after pricing questions):
- Company size: 1-10 / 11-50 / 51-200 / 200+
- Role: Founder / Manager / IC / Other
- Current solution for this problem
- How much they currently spend on alternatives per month
```

Distribute to 150-300 target prospects via email, LinkedIn, or community posts. Offer a small incentive ($5-$10 gift card) to improve response rate from ~10% to 20-30%. [src5]

**Analysis** — plot cumulative frequency curves for all 4 questions and find intersection points:

```
Manual analysis in Google Sheets:

1. Sort all responses for each question in ascending order
2. Calculate cumulative percentages at each price:
   - "Too Cheap":    % who said this price or HIGHER (descending curve)
   - "Bargain":      % who said this price or HIGHER (descending curve)
   - "Expensive":    % who said this price or LOWER  (ascending curve)
   - "Too Expensive": % who said this price or LOWER  (ascending curve)

3. Plot all 4 cumulative curves on one chart (X = price, Y = cumulative %)

4. Find intersection points:
   PMC (Point of Marginal Cheapness) = "Too Cheap" ∩ "Expensive"
     → Below this price, quality perception drops
   IPP (Indifference Price Point)    = "Bargain" ∩ "Expensive"
     → Median acceptable price; equal pull between cheap and expensive
   OPP (Optimal Price Point)         = "Too Cheap" ∩ "Too Expensive"
     → Fewest respondents reject the price at either extreme
   PME (Point of Marginal Expensiveness) = "Bargain" ∩ "Too Expensive"
     → Above this price, resistance outweighs perceived value

5. Acceptable price range = PMC to PME
   → This is where you should price. OPP is the starting recommendation.

Example output:
   PMC = $22 | IPP = $35 | OPP = $31 | PME = $58
   Acceptable range: $22 - $58
```

**Verify**: 100+ responses collected; all four curves plotted; PMC, IPP, OPP, and PME intersections identified; acceptable range is narrower than your initial test range
**If failed**: If response rate is under 10%, add incentives and extend by 1 week. If curves never intersect, sample is too small or price range tested is too narrow — recruit more respondents or ask about competitor-priced alternatives first to anchor responses. [src5]

### Step 3: Run Gabor-Granger Purchase Probability Study (Revenue Optimization)

**Duration**: 1-2 weeks (can run in parallel with Van Westendorp using separate respondent pool)
**Tool**: Google Forms (branching logic) or Conjointly

Design a sequential price test using 5-8 price points within or near the Van Westendorp acceptable range, spaced at 15-30% increments. The Gabor-Granger method starts at a middle price and branches up or down based on responses — each respondent sees only 3-5 questions, not all price points. [src6]

```
Gabor-Granger survey structure (branching):

[Same product description as Van Westendorp survey]

Start: "Would you purchase [product] at $[middle price]/month?"
├── Yes → "Would you purchase at $[higher price]?"
│   ├── Yes → "Would you purchase at $[even higher]?"
│   │   └── Record highest accepted price
│   └── No → Record previous price as highest accepted
└── No → "Would you purchase at $[lower price]?"
    ├── Yes → Record as highest accepted price
    └── No → "Would you purchase at $[even lower]?"
        └── Record response

Example price ladder (current = $29, VW range = $22-$58):
$19, $25, $29, $39, $49, $59, $69
Start at middle ($39), branch up/down.
Each respondent sees 3-5 questions due to branching.
```

**Analysis** — calculate demand curve and revenue curve:

```
For each price point, calculate:
1. % of respondents willing to buy at that price (raw)
2. Apply Newton-Miller-Smith discount factors to adjust:
   - "Definitely buy" response × 0.75
   - "Probably buy" response × 0.25
   - "Might buy" and below × 0.00
3. Adjusted purchase probability = weighted sum
4. Estimated customers = total addressable × adjusted_probability
5. Revenue at price = estimated_customers × price

Example analysis (1,000 addressable prospects):
| Price  | Raw WTP | Adjusted WTP | Est. Customers | Revenue    |
|--------|---------|-------------|----------------|------------|
| $19/mo | 82%     | 48%         | 480            | $9,120     |
| $25/mo | 71%     | 41%         | 410            | $10,250    |
| $29/mo | 63%     | 36%         | 360            | $10,440    |
| $39/mo | 44%     | 24%         | 240            | $9,360     |
| $49/mo | 28%     | 14%         | 140            | $6,860     |
| $59/mo | 15%     | 7%          | 70             | $4,130     |
| $69/mo | 7%      | 3%          | 30             | $2,070     |

Revenue-maximizing price = $29/mo (peak of adjusted revenue curve)
Note: without NMS adjustment, peak would appear at $39 — overstating by 34%
```

Plot both the demand curve (price vs. adjusted purchase probability) and the revenue curve (price x estimated customers). The peak of the revenue curve is the revenue-maximizing price. [src3]

**Verify**: 150+ responses; demand curve is monotonically downward-sloping; revenue curve has a clear single peak; peak falls within or near Van Westendorp acceptable range
**If failed**: If demand curve is flat (everyone says yes to every price), all test prices are below perceived value — extend range upward by 50-100% and re-survey. If curve drops off a cliff, perceived value may be low relative to alternatives — revisit value proposition messaging. [src6]

### Step 4: Synthesize Survey Findings and Select Test Prices

**Duration**: 1-2 days
**Tool**: Google Sheets

Overlay Van Westendorp acceptable range with Gabor-Granger revenue peak. Segment results by customer type to identify price discrimination (tiering) opportunities. Compare against competitor pricing and current price. [src4]

```
Synthesis framework:

┌──────────────────────────────────────────────────────────────────┐
│ Van Westendorp Results             Gabor-Granger Results         │
│ PMC (floor):       $22             Revenue peak (adjusted): $29  │
│ IPP (indifference): $35            Demand cliff begins:    $49   │
│ OPP (optimal):     $31             Elasticity: -1.8 at $39      │
│ PME (ceiling):     $58                                           │
│                                                                  │
│ Segment analysis:                                                │
│ - SMB (<50 employees): VW OPP = $25, GG peak = $25              │
│ - Mid-market (50-200): VW OPP = $42, GG peak = $39              │
│ - Enterprise (200+):   VW OPP = $65, GG peak = $59              │
│                                                                  │
│ Recommended actions:                                             │
│ Option 1: Single price at $29-$35 (GG peak within VW range)     │
│ Option 2: Two tiers — $29 SMB / $49 mid-market                  │
│ Option 3: Three tiers — $25 / $39 / $69                         │
│                                                                  │
│ If proceeding to live test (Path C/D), test prices:              │
│ Variant A (control):  $29 (current)                              │
│ Variant B (growth):   $39 (within VW range, strong GG signal)    │
│ Variant C (premium):  $49 (near demand cliff — risky but high)   │
└──────────────────────────────────────────────────────────────────┘
```

**Verify**: Survey data produces a clear acceptable range; GG peak falls within VW range; 2-3 test prices selected with rationale documented
**If failed**: If VW and GG results contradict (GG peak outside VW range), prioritize Gabor-Granger — it models revenue directly while Van Westendorp only maps psychological boundaries. But investigate: the contradiction often means segmentation is masking two distinct buyer groups. [src4]

### Step 5: Design and Execute Controlled Live Experiment (Path C/D Only)

**Duration**: 1 week design + 4-6 weeks run
**Tool**: Stripe/billing system, PostHog, Google Sheets

Choose a safe experiment structure that prevents brand damage. Never show different prices to the same audience on the same channel. [src7]

```
Safe experiment structures (pick one):

├── New-customer-only test (SAFEST)
│   └── Show new price to all new signups
│   └── Existing customers keep old price
│   └── No existing customer ever sees a different price
│
├── Geographic test
│   └── Different prices in different markets
│   └── Adjust for purchasing power parity
│   └── Risk: customers in different regions compare notes online
│
├── Cohort test
│   └── Alternate pricing by signup week or acquisition channel
│   └── Risk: customers from different weeks overlap in communities
│
└── Feature-gated test (SAFEST alternative)
    └── Same base price but different feature bundles per tier
    └── Tests value perception, not raw price sensitivity
    └── No customer sees a "worse deal" — different packages
```

Implementation:

```
Implementation checklist:

1. Create pricing variants in Stripe (or billing system):
   stripe prices create --currency=usd --unit-amount=2900 \
     --product=prod_XXX --nickname="control_29"
   stripe prices create --currency=usd --unit-amount=3900 \
     --product=prod_XXX --nickname="test_39"

2. Set up variant assignment:
   - Random by user ID hash (for A/B) — e.g., hash(email) % 2
   - OR time-based cohort (week 1 = $29, week 2 = $39)
   - OR geographic (US = variant A, EU = variant B)
   - Use PostHog feature flags or LaunchDarkly for assignment

3. Track per-variant (minimum 4 weeks):
   - Visitor count
   - Trial starts
   - Paid conversions
   - Conversion rate (visitor → paid)
   - Average revenue per user (ARPU)
   - Revenue per visitor (= conversion_rate × price)
   - 30-day churn rate
   - Support ticket volume (price complaints)

4. Statistical requirements:
   - Minimum 200 conversions per variant
   - Runtime: 4 weeks minimum (2 full business cycles)
   - Significance test: chi-square for conversion, t-test for RPV
   - Target: p < 0.05, power = 80%
   - Do NOT peek and stop early — inflates false positive rate

5. If a variant performs catastrophically (conversion drops 50%+),
   stop that variant but document the finding.
```

**Verify**: Experiment running with analytics tracking; variant assignment working correctly; no audience contamination between variants; 200+ conversions per variant collected over 4+ weeks
**If failed**: If fewer than 100 conversions after 4 weeks, traffic is too low for live A/B testing — rely on survey results from Steps 2-3. Revisit live testing when monthly conversions exceed 500. [src2]

### Step 6: Analyze Results and Implement Winning Price

**Duration**: 3-5 days analysis + 1 week implementation
**Tool**: Google Sheets, billing system, all customer-facing channels

Calculate per-variant metrics. Use revenue per visitor as primary metric — it captures both price and conversion impact simultaneously. A 20% price increase with a 10% conversion drop is still a net positive (+8% RPV). [src7]

```
Results analysis template:

| Metric                 | Variant A ($29) | Variant B ($39) | Variant C ($49) |
|------------------------|-----------------|-----------------|-----------------|
| Visitors               | 1,200           | 1,180           | 1,210           |
| Paid conversions       | 384 (32.0%)     | 295 (25.0%)     | 182 (15.0%)     |
| Revenue                | $11,136         | $11,505         | $8,918          |
| Revenue/visitor        | $9.28           | $9.75           | $7.37           |
| Stat sig vs control    | —               | p=0.04 ✓        | p=0.01 ✓        |
| Est. 12-month LTV      | $348            | $468            | $588            |
| 30-day churn           | 2.1%            | 2.3%            | 3.8%            |
| Support tickets (price)| 0               | 2               | 8               |

Decision: Variant B ($39) wins on revenue per visitor (+5.1%)
with acceptable churn increase (+0.2pp). LTV improvement of +34%
supports long-term value. Price complaints minimal (2 tickets).

12-month revenue projection at current traffic:
- $29 price: $133,632/year
- $39 price: $138,060/year (+3.3% revenue, +34% LTV)
```

Implementation rollout:

1. Update pricing on website, app, billing system, CRM price books, and sales materials simultaneously — all channels must show the same price. [src2]
2. Grandfather existing customers at old price for 6-12 months (or until contract renewal).
3. Create FAQ document for customer-facing teams addressing price change questions.
4. Brief sales team and CS team before pricing page goes live — a salesperson learning about a price change from a prospect is unacceptable. [src7]
5. Prepare migration communications for existing customers if they will be affected at renewal.
6. Set up 30-day and 60-day monitoring: track conversion rate, churn, support ticket volume, and net revenue impact weekly.

**Verify**: New price live across all channels simultaneously; sales team briefed before page goes live; existing customers grandfathered; 30-day monitoring dashboard active
**If failed**: If post-launch churn spikes above 5% increase vs. baseline, consider rolling back for existing customers while keeping new price for new signups only. Revisit pricing in 90 days with additional data.

## Output Schema

```json
{
  "output_type": "pricing_experiment_report",
  "format": "document + spreadsheet",
  "columns": [
    {"name": "vw_pmc", "type": "number", "description": "Van Westendorp Point of Marginal Cheapness — price floor", "required": true},
    {"name": "vw_ipp", "type": "number", "description": "Van Westendorp Indifference Price Point — median acceptable", "required": true},
    {"name": "vw_opp", "type": "number", "description": "Van Westendorp Optimal Price Point — fewest reject", "required": true},
    {"name": "vw_pme", "type": "number", "description": "Van Westendorp Point of Marginal Expensiveness — ceiling", "required": true},
    {"name": "gg_revenue_peak", "type": "number", "description": "Gabor-Granger revenue-maximizing price (NMS-adjusted)", "required": true},
    {"name": "gg_demand_curve", "type": "array", "description": "Adjusted purchase probability at each tested price point", "required": true},
    {"name": "recommended_price", "type": "number", "description": "Final recommended price based on all evidence", "required": true},
    {"name": "experiment_winner", "type": "string", "description": "Winning variant from live A/B test if conducted", "required": false},
    {"name": "revenue_per_visitor", "type": "number", "description": "Revenue per visitor for winning variant", "required": false},
    {"name": "segment_prices", "type": "object", "description": "Recommended prices by customer segment if tiered", "required": false},
    {"name": "confidence_level", "type": "string", "description": "Statistical confidence — p-value for A/B, sample size for survey", "required": true}
  ],
  "expected_row_count": "1 primary recommendation + 1 per segment (if tiered)",
  "sort_order": "primary recommendation first",
  "deduplication_key": "product + temporal_scope"
}
```

## Quality Benchmarks

| Quality Metric | Minimum Acceptable | Good | Excellent |
|---------------|-------------------|------|-----------|
| Van Westendorp responses | 100 | 150 | 200+ |
| Gabor-Granger responses | 150 | 250 | 400+ |
| Survey response rate | 10% | 20% | 30%+ |
| VW range clarity (PME/PMC ratio) | < 5x | < 3x | < 2x |
| GG revenue curve has clear peak | Yes (any) | Peak within VW range | Peak matches VW OPP |
| Live test conversions per variant | 200 | 500 | 1,000+ |
| Statistical significance (if A/B) | p < 0.10 | p < 0.05 | p < 0.01 |
| Revenue per visitor improvement | > 0% (neutral) | > 5% | > 15% |
| Methods combined | 1 | 2 | 3 (VW + GG + live) |

**If below minimum**: Re-distribute survey with $10-$25 gift card incentives and extend collection by 1-2 weeks. If live test lacks volume, rely on survey results and implement with a 30-day monitoring plan. Directional data from 100 responses is better than no data. [src5]

## Error Handling

| Error | Likely Cause | Recovery Action |
|-------|-------------|----------------|
| Survey response rate under 5% | Poor subject line, wrong audience, or no incentive | Shorten survey to essentials; add $10 gift card; distribute via LinkedIn DMs and community posts instead of mass email |
| Van Westendorp curves never intersect | Sample too small or respondents confused by questions | Collect 50+ more responses; add an example price context ("similar products cost $X-$Y"); simplify product description |
| Gabor-Granger: everyone says "yes" to all prices | All test prices below perceived value | Extend range upward by 50-100%; add 3 higher price points; re-survey fresh respondents on expanded range |
| Gabor-Granger: everyone says "no" to all prices | Product value not communicated or audience mismatch | Improve product description in survey; verify respondents match target customer profile; check if using the wrong pricing unit (per seat vs per org) |
| VW and GG results contradict each other | Different respondent pools or segment mixing | Segment both datasets by company size/role; the contradiction usually reveals two distinct buyer groups with different WTP |
| Live A/B test not reaching significance after 6 weeks | Insufficient traffic or true effect size < 5% | If effect < 5%, the difference is not commercially meaningful — keep current price and focus on value/packaging instead |
| Post-launch churn spikes > 5% above baseline | Price increase too aggressive for current customer base | Roll back to old price for existing customers; keep new price for new signups; revisit in 90 days with segment-specific pricing |
| Sales team discovers price change from prospect | Internal communication failure | Brief team immediately; create pricing FAQ doc; delay further changes until communication process is fixed |
| Survey panel responses are low quality | Respondents rushing through for incentive | Add attention check question (e.g., "Select $50 for this answer"); filter responses completed in under 60 seconds; exclude straight-line answers |

## Cost Breakdown

| Component | Free ($0) | Lean ($500) | Standard ($2K) | Advanced ($5K+) |
|-----------|-----------|-------------|----------------|-----------------|
| Survey tool | $0 (Google Forms) | $25/mo (Typeform) | $0 (Conjointly Basic) | $200/mo (Conjointly Pro) |
| Survey panel (if no email list) | $0 (DIY outreach) | $200 (social ads to survey) | $1,000 (Respondent.io / Prolific) | $3,000 (professional panel, 300 responses) |
| Response incentives | $0 | $150 ($10 x 15 responses) | $500 ($10 x 50 responses) | $1,000 ($10 x 100 responses) |
| Analysis tools | $0 (Google Sheets) | $0 (Google Sheets) | $0 (Conjointly auto-analysis) | $500 (research consultant review) |
| A/B test implementation | N/A | N/A | $500 (feature flags + engineering time) | $1,500 (full Stripe experiment setup) |
| Monitoring / analytics | $0 (PostHog free) | $0 (PostHog free) | $0 (PostHog free) | $0 (PostHog free) |
| **Total** | **$0** | **$375-$500** | **$2,000** | **$5,000-$6,200** |

## Anti-Patterns

### Wrong: Showing different prices to the same audience simultaneously
Customers who compare notes — on social media, community Slack channels, or in person — feel cheated when they discover price discrimination. Amazon famously tested variable pricing in 2000 and faced massive customer backlash. The trust damage is permanent and far exceeds any pricing insight gained. "It introduces an element of unfairness to buyers" that causes direct harm to brand reputation. [src7]

### Correct: Test on isolated audiences with zero overlap
New-customer-only tests are the safest structure. Geographic segmentation works if markets are sufficiently isolated. Time-based cohort tests (different weeks) work if your user community is not tightly knit. Feature-gated tests (different bundles at same base price) avoid the fairness issue entirely. [src2]

### Wrong: Using Van Westendorp alone to set the final price
Van Westendorp identifies a psychologically acceptable range but does not tell you where within that range revenue is maximized. A $20-$60 acceptable range still leaves a 3x decision gap. In one documented case, Van Westendorp suggested $5.75-$7.67 while Gabor-Granger identified $13.89 as revenue-maximizing for the same product — nearly 2x the upper boundary. [src4]

### Correct: Combine Van Westendorp (range) with Gabor-Granger (revenue peak)
Run both surveys. Van Westendorp finds the psychological boundaries; Gabor-Granger finds the revenue-maximizing point within those boundaries. Companies using multiple pricing research methods achieve approximately 30% higher revenue growth than those relying on a single method. [src4]

### Wrong: Taking stated willingness to pay at face value
Gabor-Granger and direct WTP questions overstate real purchase intent by 10-50%. If 60% of survey respondents say they would buy at $49, actual conversion at $49 will be 30-54%. Building a revenue model on raw survey data leads to painful revenue misses. [src8]

### Correct: Apply Newton-Miller-Smith discount factors, then validate with live data
Discount factors: 75% for "definitely buy," 25% for "probably buy," 0% for all others. This adjusts stated intent toward realistic purchase behavior. Then validate the adjusted price with a small live experiment (even 4 weeks of new-customer-only testing) before committing to a full rollout. Real behavior is the only reliable pricing signal. [src8]

## When This Matters

Use when a startup or SaaS company needs to actually execute pricing research — design the surveys, collect the data, plot the curves, calculate the revenue-maximizing point, run the experiment, and implement the validated price. Not a document about pricing theory, but the step-by-step execution with tools, sample sizes, and statistical requirements. Requires an existing product with users/prospects as input; produces a validated price point with supporting evidence as output. Most urgent before fundraising (investors scrutinize pricing power), when unit economics are failing at current prices, or when preparing to launch tiered pricing.

## Related Units

- [Idea Validation Recipe](/business/startup/idea-validation-playbook/2026) — validates demand before pricing research
- [GTM Launch Playbook](/business/startup/go-to-market-launch-playbook/2026) — validated pricing feeds pipeline targets and launch messaging
- [SaaS Pricing Benchmarks](/finance/saas-benchmarks/saas-pricing-benchmarks/2026) — calibrate experiment price ranges against industry data
- [Pricing Model Selection Framework](/business/startup/pricing-model-selection-framework/2026) — choose pricing model (seat, usage, tiered) before testing price levels
- [Financial Model Template Library](/business/startup/financial-model-template-library/2026) — pricing feeds revenue projections and fundraise modeling