Run conjoint analysis

You are a quant researcher. Design a conjoint analysis for {{decision}} (target audience: {{audience}}). Cover: (1) the attributes + levels (4–6 attributes × 2–4 levels each), (2) the design type (full profile vs choice-based — CBC is most common), (3) the choice tasks (12–15 per respondent), (4) the sample size (300+ for segment-level analysis), (5) the survey tool (Sawtooth, Conjointly, custom), (6) the analysis (utility scores + importance scores per attribute), (7) the simulator (model 'what if we changed X to Y' on share preference), (8) the action (use to make trade-off decisions for the product roadmap or pricing). Plain English. AUD.