How to Use the Choice Based Conjoint Design Prompt to Plan a Pricing Study That Answers the Question

Choice based conjoint is one of the best tools for pricing and packaging decisions, and one of the easiest to get wrong before a single respondent answers. Vague levels like "premium," attributes that overlap, a price range nobody would pay, and a sample too small for the segments you care about can all sink a study. The Choice Based Conjoint Study Designer: Attribute and Level Audit, Price Level Spacing, Tasks and Concepts per Screen, Dual Response None, Design Efficiency Test, Sample Size Rule of Thumb, HB Estimation, and Simulator Plan prompt turns a draft attribute list into a design memo that covers levels, task design, a design test, sample size, data quality, and the analysis plan.

What the prompt produces

1. A decision map that ties each attribute to a business decision and flags any that map to none.
2. An attribute and level audit for independence, clarity, the number of attributes, and the number of levels effect.
3. Price levels spaced across a believable range, with a plan for estimating price.
4. A prohibitions review that keeps impossible combinations to a minimum.
5. Task design with tasks per respondent, concepts per task, a dual response none, and holdout tasks.
6. A design efficiency test to run before fielding.
7. A sample size calculation using the Johnson and Orme rule of thumb, then adjusted for segments.
8. Data quality checks for speeders, straightliners, and fit.
9. An analysis plan covering HB utilities, importances, a share simulator, and willingness to pay by simulation.

How to fill the inputs

DecisionQuestion states what the business will decide. If an attribute does not help answer it, it probably does not belong in the study.

AttributeDraft lists your attributes and levels as you have them today, including price. Rough drafts are fine. The audit is there to fix them.

Population describes who should answer, how you screen them, and which segments you want to compare. Segments drive sample size more than anything else.

Platform names your software, such as Sawtooth Discover, Lighthouse Studio, or Conjointly, and the panel source.

SampleBudget gives the number of completes you can afford and the interview length limit.

PriceContext lists prices you see in the market so the levels stay realistic.

Reading the example output

The example designs a study for a dental clinic group considering a membership plan for patients without insurance:

• The decision map removes an attribute. "Premium service" has no definition, so the memo suggests testing a concrete feature, appointment wait time, instead.
• Levels are rewritten to be concrete, such as "2 cleanings and exams per year" rather than just "2."
• Price spans the local market with one level below and one above, and the memo explains when to switch to a linear price term.
• Task design fits the interview: ten random tasks, two holdouts, three concepts, and a dual response none question.
• The sample size math is shown. The rule of thumb gives a floor of 67, and the memo raises the plan to 400 with 200 per segment, stated as a team choice.
• Willingness to pay comes from simulation only, with a clear note that the memo contains no results.

Tips for better results

• Write levels the way a customer would read them on a menu, not the way your team talks internally.
• Run the design test before you finalize the questionnaire, and add tasks before you add respondents if a level is weak.
• Keep holdout tasks realistic so they can check whether the simulator predicts real choices.
• Plan your segment comparisons before fielding, not after.

Mistakes to avoid

• Do not include an attribute nobody can define.
• Do not set price levels far outside what the market charges.
• Do not divide utilities to get willingness to pay. Use the simulator.
• Do not report importances without explaining that they depend on the level ranges tested.

Who it is for

Market researchers and insights teams, product managers planning pricing studies, consultants scoping a conjoint project for a client, and graduate students learning study design.

Related PromptDig links

Open the Choice Based Conjoint Study Designer: Attribute and Level Audit, Price Level Spacing, Tasks and Concepts per Screen, Dual Response None, Design Efficiency Test, Sample Size Rule of Thumb, HB Estimation, and Simulator Plan prompt and paste your draft attributes to get a design memo. For more research and survey prompts, Browse more prompts. If you have a research design prompt that helped your team, Share a prompt.