# Pricing Research with Fish.Dog + Claude Code

Canonical HTML: https://fish.dog/claude-code-guide/pricing-research-guide
Markdown mirror: https://fish.dog/claude-code-guide/pricing-research-guide.md

LLM summary: Validate pricing using Van Westendorp and conjoint-adapted frameworks with 7 questions mapping price sensitivity bands and feature-tier preferences in 30 minutes.

Primary claim: Complete Claude Code guide to pricing research and willingness-to-pay analysis using Fish.Dog synthetic research. 7-question study design, Van Westendorp adaptation, feature-tier analysis, cross-segment pricing, and six deliverables from a single study.

**New to this workflow?** Read the companion article first: [How to Research Pricing with Claude Code and Fish.Dog](https://fish.dog/news/how-to-research-pricing-with-claude-code-and-ditto). It explains the concepts in plain language before you dive into the technical implementation below.

A complete Claude Code guide to pricing research and willingness-to-pay analysis. Seven-question study design adapting Van Westendorp, conjoint, and qualitative interview techniques. Feature-tier analysis, packaging preference, cross-segment pricing, and six deliverables from a single study. From guesswork to validated pricing in 30 minutes.

**For Claude Code agents and AI coding assistants.** This guide provides copy-paste API calls, study designs, and proven workflows for conducting pricing research with Fish.Dog's synthetic research API. Every command is production-tested. The output is a set of pricing recommendations backed by qualitative evidence from target personas.

Contents

1. [Why Research Pricing with Synthetic Personas](#why-pricing-research)
2. [Traditional Methods and What This Replaces](#traditional-methods)
3. [The 7-Question Pricing Study](#7-question-study)
4. [Complete API Workflow: Step by Step](#complete-api-workflow)
5. [Generating the Six Deliverables](#generating-deliverables)
6. [Deep Dive: Feature-Tier Analysis](#feature-tier-analysis)
7. [Cross-Segment Pricing Research](#cross-segment)
8. [Advanced: Testing Multiple Anchor Points](#anchoring-strategies)
9. [Worked Example: B2B Analytics SaaS](#worked-example)
10. [Connecting to Other PMM Workflows](#connecting-workflows)
11. [Best Practices and Common Mistakes](#best-practices)
12. [Frequently Asked Questions](#faq)

---

## 1. Why Research Pricing with Synthetic Personas

A 1% improvement in pricing yields an 11% improvement in operating profit (McKinsey). Yet the average SaaS company spends only six hours on pricing over the lifetime of the product (Price Intelligently / Paddle). Pricing is the most impactful revenue lever that receives the least rigorous research.

The root cause is access. Traditional pricing research is expensive, slow, and technically complex:

| Method | Time | Cost | Limitation |
| --- | --- | --- | --- |
| Conjoint analysis (Qualtrics, Sawtooth) | 4-8 weeks | $30K-$100K | Requires statistical expertise, large sample (200+), complex survey design |
| Simon-Kucher consulting | 2-4 months | $100K+ | Enterprise-only pricing; inaccessible to most companies |
| Van Westendorp survey | 2-4 weeks | $5K-$20K | Needs 200+ respondents for reliability, struggles with novel products |
| Customer interviews | 2-6 weeks | $5K-$15K | Anchoring bias, social desirability, respondents understate willingness to pay |
| **Fish.Dog + Claude Code** | **30 minutes** | **API usage only** | Qualitative directional data, not statistical precision (validated at 92% correlation) |

**The core value:** Fish.Dog pricing studies produce qualitative directional data: is your price in the right range, does your tier structure make sense, does your differentiator justify a premium. This is not a replacement for full conjoint analysis on bet-the-business decisions. It is a replacement for the informed guesswork that constitutes the pricing process at most companies.

---

## 2. Traditional Methods and What This Replaces

The 7-question study design adapts techniques from three established pricing research methods. Understanding the originals helps Claude Code agents interpret results correctly.

### Van Westendorp Price Sensitivity Meter

Developed in 1976 by Dutch economist Peter van Westendorp. Asks four questions to map an acceptable price range:

- At what price is the product **too expensive** to consider?
- At what price does it seem **so cheap** that quality must be compromised?
- At what price does it **start to feel expensive** but is still worth considering?
- At what price does it feel like **a bargain**?

The intersection points define the range of acceptable pricing and the optimal price point. Our Question 2 adapts this into a qualitative format suitable for Fish.Dog's persona model: instead of four precise price thresholds, we test gut reactions to two specific price points and ask for the walk-away threshold.

### Conjoint Analysis

Presents respondents with combinations of features and prices, forcing trade-off decisions that reveal implicit willingness to pay for each feature independently. Our Questions 4 and 5 adapt this principle: instead of statistical trade-off matrices, we ask directly which features are table-stakes versus premium, and which single feature they would keep above all others.

### Qualitative Pricing Interviews

Open-ended conversations about value, price perception, and purchase decision factors. Our Questions 1, 3, 6, and 7 draw from this tradition, capturing the emotional and contextual dimensions that quantitative methods miss: how people *feel* about what they pay, what packaging model they prefer and why, and whether a competitor's lower price would trigger switching.

**Key distinction:** Fish.Dog pricing studies do not produce a single "optimal price point" the way conjoint analysis does. They produce a *price sensitivity band* (acceptable range), *feature-tier preferences* (what goes where), and *qualitative reasoning* (why personas react the way they do). For most pricing decisions, this qualitative depth is more actionable than a statistically precise number with no context.

---

## 3. The 7-Question Pricing Study

Each question targets a specific dimension of pricing perception. Claude Code customises the bracketed sections with the actual product details, price points, and feature list before running the study.

| Q# | Question | Pricing Dimension | Adapted From |
| --- | --- | --- | --- |
| 1 | "What do you currently spend on [category] solutions per month or year? How do you feel about that price: is it fair, too much, or a bargain?" | Current spend benchmark | Qualitative interview |
| 2 | "If [product] costs [Price A], what is your gut reaction? Now what about [Price B]? At what price would you say 'that is too expensive, I am out'?" | Price sensitivity + thresholds | Van Westendorp (adapted) |
| 3 | "Would you prefer: (a) a free tier with limits, (b) a low-cost plan with core features, or (c) a premium all-in-one plan? Why?" | Packaging preference | Packaging research |
| 4 | "Which of these features would you pay extra for? Which should be included in every plan? [Feature list]" | Feature-tier allocation | Conjoint (adapted) |
| 5 | "If you could only afford one feature, which would you keep? Which feature, if removed, would make you cancel?" | Must-have vs nice-to-have | MaxDiff / conjoint |
| 6 | "Annual billing saves 20% but locks you in. Monthly billing is flexible but costs more. Which do you prefer and why?" | Billing preference | Pricing psychology |
| 7 | "A competitor offers [similar product] at [lower price] but without [key differentiator]. Would you switch? What would keep you?" | Competitive price-value trade-off | Win/loss analysis |

### Customisation Rules

**Critical:** The price points in Question 2 must be chosen carefully. Test two points that bracket your expected price: one at the lower end of your range and one at the upper end. Do not test obviously absurd prices ($1 or $10,000) as this wastes the question. The walk-away threshold captures the ceiling.

Before running the study, Claude Code should:

1. **Research the product** via web search to understand features, competitors, and current pricing (if publicly available)
2. **Identify the competitive price range** by checking 2-3 competitors' pricing pages
3. **Set anchor points** for Q2: [Price A] should be the competitive midpoint, [Price B] should be 1.5-2x the midpoint
4. **Build the feature list** for Q4: include 6-8 features covering core functionality, differentiators, and premium add-ons
5. **Identify the key differentiator** for Q7: the feature or capability that competitors lack

### Question Sequencing

Questions must be asked in order, one at a time, waiting for all persona responses before proceeding to the next question. The sequence is deliberate:

- Q1 establishes the current spend context before introducing your pricing
- Q2 tests specific prices while the current spend benchmark is fresh
- Q3-Q5 explore packaging and features before the competitive comparison
- Q6 captures billing preference as a standalone decision
- Q7 stress-tests pricing against a competitive alternative last, so it does not anchor earlier responses

---

## 4. Complete API Workflow: Step by Step

### Step 1: Create the Research Group

Recruit personas matching your target buyer profile. For pricing research, match demographics to your actual customer base as closely as possible.

```
curl -s -X POST "https://cat.fish.dog/v1/research-groups/recruit" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Pricing Study: US SaaS Buyers (30-50)",
    "description": "US professionals aged 30-50 for pricing validation of analytics SaaS",
    "group_size": 10,
    "filters": {
      "country": "US",
      "age_min": 30,
      "age_max": 50,
      "employment_status": "Employed"
    },
    "sampling_method": "random",
    "deduplicate": true
  }'
```

Save the `group.uuid` from the response.

### Step 2: Create the Research Study

```
curl -s -X POST "https://cat.fish.dog/v1/research-studies" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "title": "Pricing Validation: DataPulse Analytics Platform",
    "objective": "Validate pricing structure, willingness-to-pay thresholds, feature-tier allocation, and packaging preferences for a B2B analytics SaaS targeting marketing and product teams",
    "research_group_uuid": "GROUP_UUID_FROM_STEP_1",
    "shareable": true
  }'
```

Save the `study.id` from the response.

### Step 3: Ask Questions Sequentially

**Ask one question at a time.** Wait for all jobs to reach `"finished"` status before asking the next question. Asking multiple questions simultaneously produces unreliable responses because personas lose conversational context.

```
# Question 1: Current Spend Benchmark
curl -s -X POST "https://cat.fish.dog/v1/research-studies/STUDY_ID/questions" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "question": "What do you currently spend on analytics or business intelligence solutions per month or year? How do you feel about that price — is it fair, too much, or a bargain? What are you getting for that money?"
  }'

# Save the job_ids from the response, then poll each one:

# Poll job status (repeat until status = "finished" for ALL jobs)
curl -s "https://cat.fish.dog/v1/jobs/JOB_ID" \
  -H "Authorization: Bearer YOUR_API_KEY"

# When all jobs are finished, ask Question 2:

# Question 2: Price Sensitivity
curl -s -X POST "https://cat.fish.dog/v1/research-studies/STUDY_ID/questions" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "question": "If DataPulse costs $79 per month, what is your gut reaction? Now what about $199 per month? At what monthly price would you say that is too expensive and you would look for alternatives?"
  }'

# Wait for all jobs to finish, then Question 3:

# Question 3: Packaging Preference
curl -s -X POST "https://cat.fish.dog/v1/research-studies/STUDY_ID/questions" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "question": "Would you prefer: (a) a free tier with limited dashboards and data sources, (b) a $79/month plan with core analytics and reporting, or (c) a $199/month premium plan with AI insights, unlimited sources, and team collaboration? Why?"
  }'

# Wait, then Question 4:

# Question 4: Feature-Tier Allocation
curl -s -X POST "https://cat.fish.dog/v1/research-studies/STUDY_ID/questions" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "question": "Here are features DataPulse offers: (1) Real-time dashboards, (2) Custom reports, (3) AI anomaly detection, (4) Slack/email alerts, (5) API access, (6) Multi-team collaboration, (7) Predictive forecasting, (8) White-label client reports. Which should be included in every plan? Which would you pay extra for? Which do you not care about?"
  }'

# Wait, then Question 5:

# Question 5: Must-Have vs Cancel Trigger
curl -s -X POST "https://cat.fish.dog/v1/research-studies/STUDY_ID/questions" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "question": "If you could only afford one feature from DataPulse, which would you keep above all others? And which feature, if it were removed, would make you cancel your subscription entirely?"
  }'

# Wait, then Question 6:

# Question 6: Billing Preference
curl -s -X POST "https://cat.fish.dog/v1/research-studies/STUDY_ID/questions" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "question": "Annual billing saves 20% but locks you in for 12 months. Monthly billing is flexible but costs more. Which would you choose and why? What would make you commit to annual?"
  }'

# Wait, then Question 7:

# Question 7: Competitive Price-Value Trade-off
curl -s -X POST "https://cat.fish.dog/v1/research-studies/STUDY_ID/questions" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "question": "A competitor offers similar analytics dashboards at $49 per month but without AI anomaly detection or predictive forecasting. Would you choose the cheaper option? What would DataPulse need to prove or offer to justify the higher price?"
  }'
```

### Step 4: Complete the Study

```
# After all 7 questions have been answered:
curl -s -X POST "https://cat.fish.dog/v1/research-studies/STUDY_ID/complete" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json"

# Poll the completion jobs until finished (generates AI summary and insights)
```

### Step 5: Generate Share Link

```
curl -s -X POST "https://cat.fish.dog/v1/research-studies/STUDY_ID/share" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json"

# Returns: { "share_url": "https://cat.fish.dog/organization/studies/shared/TOKEN" }
```

### Step 6: Fetch All Responses for Analysis

```
curl -s "https://cat.fish.dog/v1/research-studies/STUDY_ID/questions" \
  -H "Authorization: Bearer YOUR_API_KEY"

# Returns all questions with all persona responses
# Use this data to generate the six deliverables
```

---

## 5. Generating the Six Deliverables

A completed pricing study yields 70 qualitative responses (10 personas x 7 questions). Claude Code analyses these and produces six structured deliverables:

| Deliverable | Source Questions | What It Contains | Who Uses It |
| --- | --- | --- | --- |
| **Price Sensitivity Band** | Q1, Q2 | Acceptable price range derived from persona reactions and stated walk-away thresholds. Not a single number but a directional range with qualitative reasoning. | Founders, PMM, Finance |
| **Feature-Tier Recommendation** | Q4, Q5 | Proposed tier structure: which features are base (table-stakes), which are premium upsell candidates, which generated insufficient interest for any tier. | Product, PMM, Sales |
| **Packaging Preference Breakdown** | Q3 | Split between freemium, flat-rate, and tiered preferences with the qualitative reasoning behind each. Directly informs the packaging model decision. | Product, PMM, Growth |
| **Price Anchoring Strategy** | Q1, Q2 | How pricing should be framed relative to current spend and perceived value. Includes emotional context: "painful but necessary" vs "fair" vs "bargain". | Marketing, Sales, Web |
| **Competitive Price Positioning** | Q7 | Whether the differentiator justifies the premium over competitors. Lists what proof points, guarantees, or value would defend the price in sales conversations. | Sales, PMM, CS |
| **Willingness-to-Pay by Persona** | Q2, Q7, demographics | Price sensitivity variation by persona profile. Maps demographic/psychographic differences to different thresholds and preferences. | PMM, Sales, Product |

### Generating the Price Sensitivity Band

Parse Q2 responses to extract three data points per persona: reaction to Price A, reaction to Price B, and stated walk-away threshold.

```
# Pseudocode for price sensitivity band extraction
responses_q1 = fetch_responses(study_id, question_1_id)
responses_q2 = fetch_responses(study_id, question_2_id)

current_spend = []
price_a_reactions = []  # e.g., "fair", "cheap", "expensive", "about right"
price_b_reactions = []
walk_away_thresholds = []

for response in responses_q2:
    text = response["response_text"]
    # Extract: reaction to Price A (positive/neutral/negative)
    # Extract: reaction to Price B (positive/neutral/negative)
    # Extract: stated walk-away threshold (dollar amount)
    # Cross-reference with Q1 current spend for context

# Output:
# PRICE SENSITIVITY BAND
# =======================
# Current spend range: $50-$300/month (median: $120)
# Emotional context: 6/10 describe current spend as "fair", 3/10 as "too much"
#
# Price A ($79/mo) reactions:
#   7/10 positive ("reasonable", "about right", "good value")
#   2/10 neutral ("depends on features")
#   1/10 negative ("still need to justify internally")
#
# Price B ($199/mo) reactions:
#   3/10 positive ("worth it if AI features deliver")
#   4/10 conditional ("need to see ROI first")
#   3/10 negative ("too expensive for my team size")
#
# Walk-away thresholds: range $150-$500, median $250
#
# RECOMMENDED RANGE: $79-$199/month
# Sweet spot: $99-$149 (positive reactions from 8/10, below median walk-away)
```

### Generating the Feature-Tier Recommendation

```
# Parse Q4 (feature allocation) and Q5 (must-have/cancel trigger)
responses_q4 = fetch_responses(study_id, question_4_id)
responses_q5 = fetch_responses(study_id, question_5_id)

# For each feature, count:
# - "should be in every plan" mentions (table-stakes)
# - "would pay extra" mentions (premium)
# - "don't care" mentions (deprioritise)

# Cross-reference with Q5:
# - "would keep above all others" = core value driver
# - "removal would make me cancel" = retention anchor

# Output:
# FEATURE-TIER RECOMMENDATION
# ============================
#
# BASE TIER (include in all plans):
#   - Real-time dashboards        [9/10 "every plan", Q5: #1 "would keep"]
#   - Custom reports              [8/10 "every plan"]
#   - Slack/email alerts           [7/10 "every plan"]
#
# PREMIUM TIER (upsell candidates):
#   - AI anomaly detection         [7/10 "pay extra", Q5: #1 "cancel trigger"]
#   - Predictive forecasting       [6/10 "pay extra"]
#   - Multi-team collaboration     [5/10 "pay extra"]
#
# ENTERPRISE / ADD-ON:
#   - API access                   [4/10 "pay extra", 3/10 "don't care"]
#   - White-label client reports   [3/10 "pay extra", 5/10 "don't care"]
#
# KEY INSIGHT: AI anomaly detection is the #1 cancel trigger (retention
# anchor) but NOT the #1 "would keep" feature (acquisition driver).
# This means: include it in the base tier to retain, but position
# real-time dashboards as the hero feature to acquire.
```

### Generating the Packaging Preference Breakdown

```
# Parse Q3 responses for preference and reasoning
responses_q3 = fetch_responses(study_id, question_3_id)

preferences = {"free_tier": 0, "low_cost": 0, "premium": 0}
reasoning = {"free_tier": [], "low_cost": [], "premium": []}

for response in responses_q3:
    # Identify which option the persona chose
    # Extract their reasoning
    # Categorise the reasoning (e.g., "need to prove value", "hate limits",
    #   "want everything", "budget constraints")

# Output:
# PACKAGING PREFERENCE BREAKDOWN
# ================================
# Free tier:    3/10 (30%)
#   Reasons: "Need to prove value to my boss before requesting budget" (2x),
#            "Want to explore before committing" (1x)
#
# Low-cost ($79): 5/10 (50%)
#   Reasons: "Core analytics is all I need" (3x),
#            "Premium feels like overkill for a small team" (2x)
#
# Premium ($199): 2/10 (20%)
#   Reasons: "AI insights are the whole point" (1x),
#            "Hate being nickel-and-dimed on upgrades" (1x)
#
# IMPLICATION: The market wants a free-to-paid conversion path.
# Lead with a free tier for acquisition, price core analytics
# at $79, and position AI features as the premium upgrade.
```

---

## 6. Deep Dive: Feature-Tier Analysis

Feature-tier allocation is the most directly actionable output of a pricing study. It answers the question most pricing pages get wrong: what goes where.

### The Three Categories

| Category | Signal from Personas | Pricing Implication |
| --- | --- | --- |
| **Table-stakes** | 7+ of 10 personas say "should be in every plan" | Include in the base tier. Gating these behind a paywall creates friction without premium perception. |
| **Premium** | 5+ of 10 say "would pay extra", fewer than 5 say "every plan" | Gate behind a higher tier. These features carry pricing power and justify the upgrade. |
| **Deprioritise** | 4+ of 10 say "don't care" or don't mention it | Consider removing from the pricing page entirely. Including features nobody values clutters the comparison and dilutes perceived value of features people do care about. |

### Acquisition vs Retention Features

Question 5 reveals a critical distinction that most pricing analyses miss:

- **Acquisition driver** (Q5: "would keep above all others"): the feature that makes people sign up. This should be prominent in your base tier and central to your messaging.
- **Retention anchor** (Q5: "removal would make me cancel"): the feature that makes people stay. This should be available early enough in the customer journey that switching costs accumulate.

**When these differ, you have a strategic insight.** If the acquisition driver is "real-time dashboards" but the retention anchor is "AI anomaly detection," your product strategy should include AI features in every tier (even a limited version) so that retention value starts building immediately, even though dashboards are the reason people showed up.

### Interpreting Mixed Signals

Some features will receive split responses: 4/10 say "every plan," 4/10 say "pay extra," 2/10 say "don't care." This split usually means the feature's value depends on the persona's context. Cross-reference with demographics:

- If the "every plan" responses come from senior personas and "pay extra" from junior ones, the feature is a table-stakes expectation at the enterprise level
- If the split correlates with team size, company stage, or technical sophistication, you may need different tiers for different segments rather than a one-size-fits-all pricing page

---

## 7. Cross-Segment Pricing Research

The basic study tests pricing with one audience. Cross-segment pricing reveals how price sensitivity varies by buyer type, which is where most pricing decisions get complex.

### Running Parallel Studies

Claude Code orchestrates three studies concurrently with different persona groups, all using the same seven questions:

```
# Group 1: SMB decision-makers
curl -s -X POST "https://cat.fish.dog/v1/research-groups/recruit" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Pricing Study: SMB Buyers (25-40)",
    "group_size": 10,
    "filters": {
      "country": "US",
      "age_min": 25,
      "age_max": 40,
      "employment_status": "Employed"
    }
  }'

# Group 2: Enterprise buyers
curl -s -X POST "https://cat.fish.dog/v1/research-groups/recruit" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Pricing Study: Enterprise Buyers (35-55)",
    "group_size": 10,
    "filters": {
      "country": "US",
      "age_min": 35,
      "age_max": 55,
      "employment_status": "Employed"
    }
  }'

# Group 3: Technical evaluators
curl -s -X POST "https://cat.fish.dog/v1/research-groups/recruit" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Pricing Study: Technical Buyers (Bachelor+)",
    "group_size": 10,
    "filters": {
      "country": "US",
      "age_min": 25,
      "age_max": 50,
      "employment_status": "Employed",
      "education_level": "Bachelors"
    }
  }'
```

### Cross-Segment Analysis Output

```
CROSS-SEGMENT PRICING MATRIX
==============================

                    SMB (25-40)      Enterprise (35-55)   Technical (25-50)
Current spend:      $30-$100/mo      $200-$800/mo         $50-$200/mo
Price A ($79):      8/10 positive    6/10 "cheap?"        7/10 positive
Price B ($199):     2/10 positive    8/10 positive        5/10 conditional
Walk-away:          median $120      median $500           median $250
Packaging pref:     Free tier (6/10) Premium (7/10)       Low-cost (5/10)
Billing pref:       Monthly (8/10)   Annual (7/10)        Monthly (6/10)
Competitor switch:  6/10 would       2/10 would           4/10 would

IMPLICATIONS:
- SMB needs a free-to-$79 path with monthly billing and fast time-to-value
- Enterprise expects $199+ pricing and prefers annual contracts with sales
- Technical buyers are price-moderate but need capability proof before committing
- A single pricing page with 3 tiers can work, BUT the sales motion
  must differ: self-serve for SMB, sales-led for enterprise
- The $79 price point risks "too cheap" perception for enterprise buyers
```

**GTM implication:** If your SMB segment strongly prefers free-tier entry and your enterprise segment expects a sales conversation, you need a hybrid motion (product-led sales). The pricing study does not just inform the numbers. It informs how you sell.

---

## 8. Advanced: Testing Multiple Anchor Points

Question 2 presents specific price points, which means persona responses are anchored to those numbers. This is both useful (tests specific prices) and limiting (anchoring bias). The mitigation: run two parallel studies with different anchor points.

### Dual-Anchor Study Design

```
# Study A: Low anchor
# Q2: "If DataPulse costs $49/mo, what is your gut reaction?
#      Now what about $129/mo? At what price is it too expensive?"

# Study B: High anchor
# Q2: "If DataPulse costs $99/mo, what is your gut reaction?
#      Now what about $249/mo? At what price is it too expensive?"

# All other questions remain identical.
# Run both studies concurrently with similar persona groups.
```

### Interpreting Dual-Anchor Results

| Signal | Interpretation | Action |
| --- | --- | --- |
| Walk-away thresholds are similar across both studies | The ceiling is real, not an artefact of anchoring | Price below the shared ceiling with confidence |
| Walk-away is higher in Study B (high anchor) | Anchoring effect: personas adjusted expectations upward | The true ceiling is between the two; lean toward the higher Study B data |
| Positive reactions to $129 (Study A) but negative to $99 (Study B) | Framing effect: $129 feels like a deal when compared to $49, but $99 feels expensive when compared to $249 | Your pricing page framing matters as much as the number itself. Test pricing page layouts, not just prices. |

**Time cost:** Dual-anchor testing doubles the study time to ~60 minutes. Only use this approach when the price point is high-stakes (e.g., repricing an existing product with a large customer base). For new products, a single study with well-chosen anchors is sufficient.

---

## 9. Worked Example: B2B Analytics SaaS

### Context: DataPulse Analytics

**Product:** B2B analytics platform for marketing and product teams. Real-time dashboards, custom reports, AI anomaly detection, predictive forecasting.

**Current state:** Pricing set at launch 18 months ago by looking at Mixpanel, Amplitude, and Heap. Currently $99/month flat rate, no tiers. Founder suspects they are underpriced but is nervous about raising prices.

**Goal:** Validate whether a tiered pricing model ($79 / $149 / $249) would be accepted, and which features belong in each tier.

### Study Setup

```
Group: 10 US adults aged 30-50, employed (matching ICP: marketing managers,
       product leads, data analysts)
Study: "DataPulse Pricing Validation"
Anchor points: $79 and $199 (Q2)
Feature list: Real-time dashboards, Custom reports, AI anomaly detection,
              Slack/email alerts, API access, Multi-team collaboration,
              Predictive forecasting, White-label client reports
Competitor for Q7: "A competitor offers dashboards and reports at $49/mo
                    but without AI features"
```

### Results Summary

```
Q1 - CURRENT SPEND:
  Range: $0 (using free tools) to $400/mo
  Median: $120/mo
  Emotional context: 5/10 "fair", 3/10 "too much for what I get", 2/10 "bargain"
  Key quote: "I pay $200/mo for Amplitude and use maybe 30% of the features.
  I'd pay the same for something simpler that I actually use." - Maria, 38,
  Marketing Director, Chicago

Q2 - PRICE SENSITIVITY:
  $79/mo: 8/10 positive ("reasonable", "good value", "I'd try it")
  $199/mo: 4/10 positive, 3/10 conditional ("need to see AI in action first"),
           3/10 negative ("hard to justify for a team of 3")
  Walk-away thresholds: $150, $200, $200, $250, $250, $300, $300, $350, $400, $500
  Median walk-away: $275

Q3 - PACKAGING:
  Free tier: 2/10 ("always want to try before I buy")
  $79 core plan: 5/10 ("just want dashboards and reports, don't need AI")
  $199 premium: 3/10 ("if the AI actually works, it's worth it")

Q4 - FEATURE TIERS:
  Every plan: Real-time dashboards (10/10), Custom reports (9/10),
              Slack alerts (7/10)
  Pay extra: AI anomaly detection (8/10), Predictive forecasting (7/10),
             Multi-team collab (5/10)
  Don't care: White-label reports (6/10 don't care), API access (4/10 don't care)

Q5 - MUST-HAVE vs CANCEL:
  Keep one feature: Real-time dashboards (6/10), AI anomaly detection (3/10),
                    Custom reports (1/10)
  Cancel trigger: AI anomaly detection (4/10), Real-time dashboards (3/10),
                  Custom reports (2/10), Predictive forecasting (1/10)

Q6 - BILLING:
  Monthly: 6/10 ("flexibility", "want to cancel if we outgrow it")
  Annual: 4/10 ("20% savings is real money", "already committed to analytics")
  Key insight: "I'd go annual if there was a 3-month trial period first" (3 personas)

Q7 - COMPETITIVE:
  Would switch to cheaper: 3/10 ("dashboards are dashboards")
  Would stay with DataPulse: 5/10 ("AI features are why I'm here")
  Conditional: 2/10 ("stay if DataPulse can prove ROI of AI features")
  Key defence: "Show me one example where the anomaly detection caught
  something I would have missed" - James, 42, Product Lead, Austin
```

### Deliverables Generated

```
PRICE SENSITIVITY BAND
=======================
Acceptable range: $79-$250/month
Sweet spot: $99-$149/month
  - $79 universally positive but may leave money on the table
  - $199 viable for premium tier but requires proof of AI value
  - Median walk-away ($275) gives headroom above current flat rate ($99)
Recommendation: Raise from $99 flat to a tiered model. Base at $79
(lower entry point increases trial conversion), Pro at $149 (captures
the sweet spot), Enterprise at $249 (just below median walk-away).

FEATURE-TIER RECOMMENDATION
============================
Starter ($79/mo):
  - Real-time dashboards
  - Custom reports (up to 10)
  - Slack/email alerts
  - Single-user

Pro ($149/mo):
  - Everything in Starter
  - AI anomaly detection
  - Predictive forecasting
  - Multi-team collaboration (up to 5 seats)

Enterprise ($249/mo):
  - Everything in Pro
  - API access
  - White-label client reports
  - Unlimited seats
  - Dedicated support

PACKAGING PREFERENCE: Freemium path (free tier with 1 dashboard,
no AI) for acquisition. 50% of personas preferred the core plan,
suggesting $79 is the volume tier.

PRICE ANCHORING STRATEGY: Frame against current spend ("most teams
spend $120-$200/mo on analytics they barely use") rather than against
competitors. Personas who described current spend as "too much for
what I get" are the primary acquisition target.

COMPETITIVE PRICE POSITIONING: AI features justify a 2x premium over
basic dashboard competitors. But the justification requires proof:
"show me one example" was the most common defence requirement. Build
a demo/case study showing anomaly detection catching a real issue.

WILLINGNESS-TO-PAY BY PERSONA:
  - Marketing directors (senior): comfortable at $199, walk away at $350
  - Product managers (mid): sweet spot at $99-$149, walk away at $250
  - Data analysts (technical): price-sensitive ($79 preferred), but
    highest feature engagement with AI tools
```

### Actionable Outcome

The founder's instinct was correct: $99 flat rate is below the market's willingness to pay. The study recommends a three-tier model ($79 / $149 / $249) with a free tier for acquisition. The critical insight: AI anomaly detection is the retention anchor (cancel trigger) but real-time dashboards are the acquisition driver (hero feature). Price the dashboards accessibly, gate the AI behind Pro, and build proof of AI value to defend the premium.

---

## 10. Connecting to Other PMM Workflows

Pricing research is downstream of positioning and messaging, but upstream of GTM execution. Each Fish.Dog study type feeds into the next:

| Sequence | Study Type | What It Produces for Pricing |
| --- | --- | --- |
| 1. Before pricing | [Positioning Validation](/claude-code-guide/validate-product-positioning-guide) | Confirms market category, competitive alternatives, and unique attributes. Your price must be congruent with your positioning: premium positioning requires premium pricing. |
| 2. Before pricing | [Messaging Testing](/claude-code-guide/test-product-messaging-guide) | Validates how to communicate value. Your pricing page *is* messaging: tier names, feature comparisons, and price anchoring are all messaging decisions. |
| 3. This guide | **Pricing Research** | Validates what the market will pay, how it wants to buy, and which features justify the premium. |
| 4. After pricing | [Competitive Intelligence](/claude-code-guide/competitive-intelligence-guide) | Provides context for pricing conversations: where competitors price, what premium your differentiation commands. |
| 5. After pricing | GTM Execution | Takes the pricing model and distributes it: self-serve pricing page, sales negotiation playbook, expansion revenue strategy. |

**Full stack timing:** Positioning validation (30 min) + Messaging testing with 2 rounds (70 min) + Pricing research (30 min) + Competitive intelligence (45 min) = under 4 hours for the complete PMM strategic foundation. This typically takes a quarter to build through traditional methods.

### Using Pricing Data in Other Deliverables

- **Sales battlecards:** The competitive price-value trade-off data (Q7) feeds directly into the "Quick Dismisses" section when competitors undercut on price
- **Messaging hierarchy:** Price anchoring strategy informs how to frame value on the pricing page, in sales decks, and in email campaigns
- **Product roadmap:** Feature-tier analysis shows which features carry pricing power (invest more) and which are indifferent (deprioritise)
- **Investor decks:** Willingness-to-pay data demonstrates pricing headroom and supports revenue projections

---

## 11. Best Practices and Common Mistakes

### Best Practices

| Practice | Why It Matters |
| --- | --- |
| **Research competitive pricing before designing the study** | Your anchor points (Q2) and competitor reference (Q7) must be realistic. Fictional competitors or implausible prices produce unreliable responses. |
| **Match persona demographics to actual buyers** | Pricing sensitivity varies enormously by age, seniority, and company size. A study with 25-year-olds will produce different thresholds than one with 45-year-olds. |
| **Include 6-8 features in Q4, not 3-4 or 15** | Too few features produces obvious answers ("all should be included"). Too many overwhelms personas and produces noisy data. |
| **Test prices that bracket your expected range** | If you expect to price at $99, test $79 and $199 (Q2). Testing $49 and $499 wastes the question on prices you would never charge. |
| **Pay attention to emotional language in Q1** | "Fair" vs "painful" vs "bargain" tells you more about pricing power than the dollar amount. A market that describes current spend as "painful" is ready for a value-framed alternative. |
| **Cross-reference Q4 and Q5** | Features that are "every plan" in Q4 but "cancel trigger" in Q5 must stay in the base tier. Never gate a retention anchor behind premium. |

### Common Mistakes

| Mistake | Why It Fails | What to Do Instead |
| --- | --- | --- |
| Testing only one price point | Gives a thumbs-up/thumbs-down with no range data | Always test two price points plus the walk-away threshold (Q2 design) |
| Skipping Q1 (current spend) | Without a benchmark, price reactions have no context | Q1 grounds every subsequent response in reality |
| Using absurd competitor prices in Q7 | "A competitor at $5/mo" is not a realistic test of switching | Use an actual competitor's price or a realistic market alternative |
| Ignoring the "why" in Q3 | Knowing 5/10 prefer the core plan is less useful than knowing *why* | The reasoning reveals GTM implications: "need to prove value" = PLG, "hate limits" = flat-rate |
| Treating the price sensitivity band as a single number | Qualitative pricing research produces a range, not an optimum | Present a range with reasoning, then use A/B pricing tests to find the exact point |
| Not customising questions for the product | Generic questions ("how much would you pay for this?") produce generic answers | Include specific product name, features, competitor names, and realistic price points |
| Running pricing study before positioning | Price is a positioning signal. Without validated positioning, you are pricing in a vacuum. | Run [positioning validation](/claude-code-guide/validate-product-positioning-guide) first, then pricing research |

---

## 12. Frequently Asked Questions

### Is synthetic pricing research reliable?

Fish.Dog's synthetic personas are validated at 92% correlation with traditional research (EY, Harvard, Cambridge, Stanford, Oxford). For pricing, the correlation is strongest for *attitudinal* data (how people feel about prices, what they perceive as fair/expensive) and directional for *behavioural* data (what they would actually pay). The recommended approach: use Fish.Dog for the fast first pass to narrow the range and validate the tier structure, then confirm with real-world A/B pricing tests.

### How accurate are the walk-away thresholds?

Walk-away thresholds from synthetic personas should be treated as directional upper bounds. Real-world walk-away thresholds may be lower (due to competitive alternatives not mentioned in the study) or higher (due to switching costs and inertia). The median walk-away across 10 personas gives you a ceiling that would be risky to exceed without exceptional justification.

### Should I use this instead of conjoint analysis?

Not for high-stakes pricing restructuring at a large company. Conjoint produces statistically precise willingness-to-pay data for individual features, which is essential when optimising pricing across multiple product lines. Use Fish.Dog for: validating pricing before launch, testing a pricing change before announcing it, exploring tier structures, understanding packaging preferences, and any situation where qualitative depth matters more than statistical precision.

### How do I handle international pricing?

Run separate studies for each market using Fish.Dog's country filters. Pricing expectations vary significantly by country. A $99/month product in the US may be perceived very differently in the UK, Germany, or Canada. Claude Code can orchestrate parallel studies across 4+ countries in the same session, producing a cross-market pricing matrix.

```
# Example: International pricing comparison
# Run the same 7 questions against groups in 4 markets:
# US: 10 personas, age 30-50, employed
# UK: 10 personas, age 30-50, employed
# Germany: 10 personas, age 30-50, employed
# Canada: 10 personas, age 30-50, employed
# Total time: ~45 minutes (studies run sequentially per market)
```

### What if my product has no competitors?

Modify Q7 to test against the status quo rather than a named competitor: "Currently, most teams solve this with spreadsheets, manual processes, or by hiring an analyst. These cost roughly [X] per month in time and salary. Would you prefer the manual approach at that cost, or DataPulse at [price]?" This tests pricing power against the "do nothing" alternative, which is the real competitor for novel products.

### How often should I run pricing research?

At minimum: before launch, before any pricing change, and when entering a new market segment. Ideally: annually as a check on whether the market has shifted. At 30 minutes per study, the time cost is negligible compared to the revenue impact of mispriced products.

### Can I test usage-based pricing?

Yes, by adapting Q3: "Would you prefer: (a) a flat monthly fee for unlimited usage, (b) a lower base fee plus per-event charges, or (c) a pay-as-you-go model with no monthly commitment? Why?" The same framework applies. The "why" reveals whether personas value predictability (flat fee), cost control (per-event), or flexibility (pay-as-you-go).

### How do I test pricing for a freemium model specifically?

Adapt Q3 to present specific freemium options: "Would you prefer: (a) a free plan with [specific limits], (b) a $X plan with [expanded limits], or (c) a $Y plan with no limits? What would make you upgrade from free to paid?" The upgrade trigger question is critical: it tells you which limit on the free tier creates the conversion event.

### What is the minimum viable pricing study?

If time is extremely constrained, a 3-question study captures the most important data:

- **Q1 (Current Spend + Q2 Price Sensitivity combined):** "What do you currently spend on [category], and if [product] cost [Price A], how would you react? At what price is it too expensive?"
- **Q4 (Feature-Tier):** "Which of these features should be included in every plan, and which would you pay extra for?"
- **Q7 (Competitive):** "A competitor offers [X] at [lower price] but without [differentiator]. Would you switch?"

This produces a price sensitivity range, feature-tier allocation, and competitive pricing defence in approximately 15 minutes. Use the full 7-question study when time allows.

---

**Related guides:**

- [Fish.Dog API for Claude Code: Complete Guide](https://fish.dog/claude-code-guide)
- [Positioning Validation Guide](https://fish.dog/claude-code-guide/validate-product-positioning-guide)
- [Competitive Intelligence Guide](https://fish.dog/claude-code-guide/competitive-intelligence-guide)
- [Product Messaging Guide](https://fish.dog/claude-code-guide/test-product-messaging-guide)
- [Voice of Customer Guide](https://fish.dog/claude-code-guide/voice-of-customer-guide)
- [Customer Segmentation Guide](https://fish.dog/claude-code-guide/customer-segmentation-guide)
- [GTM Strategy Validation Guide](https://fish.dog/claude-code-guide/gtm-strategy-validation-guide)
- [Content Marketing Engine Guide](https://fish.dog/claude-code-guide/content-marketing-engine-guide)
- [Sales Enablement Guide](https://fish.dog/claude-code-guide/sales-enablement-guide)
- [Product Launch Research Guide](https://fish.dog/claude-code-guide/product-launch-research-guide)
- [Question Design Playbook](https://fish.dog/claude-code-guide/question-design-playbook)
- [Installing the Fish.Dog Skill](https://fish.dog/claude-code/installing-ditto-skill)

**Read the articles:** *These guides are written for Claude Code agents. The articles below are written for you.*

- [How to Validate Product Positioning with Claude Code and Fish.Dog](https://fish.dog/news/how-to-validate-product-positioning-with-claude-code-and-ditto)
- [How to Build Competitive Battlecards with Claude Code and Fish.Dog](https://fish.dog/news/how-to-build-competitive-battlecards-with-claude-code-and-ditto)
- **How to Research Pricing with Claude Code and Fish.Dog** (this guide's companion article)
- [How to Test Product Messaging with Claude Code and Fish.Dog](https://fish.dog/news/how-to-test-product-messaging-with-claude-code-and-ditto)
- [How to Run Voice of Customer Research with Claude Code and Fish.Dog](https://fish.dog/news/how-to-run-voice-of-customer-research-with-claude-code-and-ditto)
- [How to Segment Customers with Claude Code and Fish.Dog](https://fish.dog/news/how-to-segment-customers-with-claude-code-and-ditto)
- [How to Validate GTM Strategy with Claude Code and Fish.Dog](https://fish.dog/news/how-to-validate-gtm-strategy-with-claude-code-and-ditto)
- [How to Build a Content Marketing Engine with Claude Code and Fish.Dog](https://fish.dog/news/how-to-build-a-content-marketing-engine-with-claude-code-and-ditto)
- [How to Build Sales Enablement with Claude Code and Fish.Dog](https://fish.dog/news/how-to-build-sales-enablement-with-claude-code-and-ditto)
- [How to Research a Product Launch with Claude Code and Fish.Dog](https://fish.dog/news/how-to-research-a-product-launch-with-claude-code-and-ditto)
- [Overview: Using Fish.Dog and Claude Code for Product Marketing](https://fish.dog/news/using-ditto-and-claude-code-for-product-marketing)