Base size refers to the number of participants included in your study or analysis. Choosing the right base size helps ensure your findings are credible, interpretable, and appropriate for your research goals.
This article explains:
What a base size is
How recruitment affects base size
How many participants to use for different study types
How to think about base sizes for concept testing
When statistical significance is relevant
What is a base size?
Your base size is the number of participants contributing data to a result, insight, or metric.
You need at least 5 interviews to begin generating insights.
The right base size depends entirely on what you’re trying to learn.
How to think about your base size and recruitment
Your ability to reach a given base size depends heavily on how easy the audience is to recruit.
Niche or low-incidence audiences
Higher base sizes may be difficult or impractical.
Examples:
Ages 18–24 living in the US; people who purchased a car in Ohio; customers who bought a Kia at a specific dealership
In these cases, smaller base sizes may still be appropriate especially for qualitative research.
Broad or high-incidence audiences
Larger base sizes are usually easier to achieve.
Example:
Past 3‑month candy eaters (non‑chocolate)
Broad audiences make a study more feasible and efficient to to field.
What influences sample size?
Several factors affect how large your base size should be, including:
Research objectives
Audience incidence (IR)
Number of segments
How you want to cut the data
How many participants do Outset studies usually have?
It depends entirely on the study type and research objective.
Across Outset, we see studies ranging from 5 to 1,000+ participants, but most commonly we see 20–50 participants, depending on audience incidence and methodology.
Sample size guidelines
Use the guidelines below to choose a base size based on what you want to learn.
Outset Qualitative Research
Where the interview has a majority of conversational questions, less quantitative questions.
Research purpose | Typical base size | What this supports |
Early exploration | 20+ total participants | Initial themes, language, and hypothesis generation |
Theme saturation | 30+ total participants | Core themes start to recur |
Pattern confidence | 40-50 total participants | Emerging confidence in relative prominence and consistency of themes |
How to choose: Use theme saturation when you want to know what themes exist. Use pattern confidence when you want to understand which themes appear more often than others.
Outset Quantitative research (directional)
Where the interview has a majority of quantitative questions (Rating, Stack Rank, Multiple Choice).
Research purpose | Recommended base size | What this supports |
Directional quantification – standard AI testing & usability (NOT concept testing) | 20–30+ total participants | Directional interpretation of quantitative metrics |
Directional quantification – concept testing | 10–20 per concept | Directional comparison of concepts without statistical certainty |
Outset Quantitative research (statistical significance)
Where the interview has a majority of quantitative questions (Rating, Stack Rank, Multiple Choice).
Research purpose | Recommended base size | What this supports |
Statistical quantification of quant metrics only (NOT concept testing) | 50+ total participants *recommended if audience is broad (high incidence) but not if audience is niche (low incidence) | Statistically significant data |
Statistical quantification of quant metrics only – concept tests | 50+ per concept *recommended if audience is broad (high incidence) but not if audience is niche (low incidence) | Statistically grounded comparison of concepts |
How should I think about base sizes for concept testing?
Concept testing base size depends on:
The number of concepts being tested
Whether concepts are tested monadically or sequentially
Key definitions
Monadic testing: Each participant sees only one concept
Sequential testing: Participants see multiple concepts in sequence
What do I need for statistical significance / analysis?
Statistical analysis typically applies to quantitative questions / analysis.
Statistical analysis does not typically apply to qualitative themes or open-ended responses.
As a rule of thumb:
Plan for 50+ total participants (or 50+ per concept) for basic statistical significance / analysis
Final note
These guidelines are intended to support planning and interpretation, not replace internal research standards. We recommend aligning base size decisions with your internal practices and research goals whenever possible.
These guidelines are meant to support planning and interpretation, not replace your internal standards. Always align base size decisions with your research goals and audience constraints.
Hope this helps! If you have any further questions, please reach out to our team at [email protected] or via chat.