Conjoint Sample Size Calculator
Calculate the minimum sample size for CBC conjoint studies using Orme's rule of thumb. Enter your design parameters to get a recommendation.
Study Design Parameters
The attribute with the most options (e.g. price with 5 levels = 5)
Number of choice sets shown to each person (typically 8–16)
Number of product options shown per task, excluding 'none' (typically 2–4)
Minimum Sample
56
Statistical minimum (Orme's formula)
Recommended Sample
56
At standard confidence level
With Recruitment Buffer
65
+15% for dropouts / disqualifications
Formula used
n ≥ (500 × max_levels) / (tasks × alternatives), per Orme (2010). For HB estimation, minimum can be lower; for aggregate logit use the conservative estimate.
Typical design benchmarks
| Use case | Tasks | Alternatives | Typical n |
|---|---|---|---|
| Simple product (3–4 attributes) | 8–10 | 2–3 | 150–250 |
| Standard CBC study | 10–14 | 3 | 250–400 |
| Segment-level analysis | 12–16 | 3–4 | 400–600 |
| Complex design (6+ attributes) | 14–20 | 3–4 | 500–800 |
Want the methodology behind this?
Read our full guide on conjoint analysis, covering study design, interpretation, and common mistakes.
Read the guide