Skip to content
Consumr.AI Onboarding

Run Quantitative Research

Quantitative research on the platform is powered by respondents: the lighter, mini-twins that represent the statistical distribution of your segment at scale. Respondents answer the “how many” and “which” questions through surveys.

For the full menu of survey types (brand track, segmentation, media consumption, custom, quick), see Methods.

When to use quantitative research

Section titled “When to use quantitative research”
  • You need to measure the prevalence of an attitude, preference, or behaviour across a segment.
  • You want to compare how multiple segments or brands perform on specific dimensions.
  • You need a brand funnel (awareness → consideration → preference → intent → endorsement) for competitive benchmarking.
  • You have a hypothesis from qualitative research and want to validate it at scale.

The procedure

Section titled “The procedure”

1. Ensure you have segments and respondents ready

Section titled “1. Ensure you have segments and respondents ready”

Quantitative surveys run on respondents. If you have built AI twins using the Segments path, you can create respondents from those twins. If you have not yet created segments, see Create Segments — the 4 Methods.

2. Navigate to the survey research area

Section titled “2. Navigate to the survey research area”

Open the platform and navigate to the quantitative (Quant / Survey) section. You will see two main tracks:

  • Standard surveys: pre-built survey types with locked, validated question sets. These cover brand track, segmentation, and media consumption.
  • Custom and quick surveys: you build your own question set or run a poll.

3. Select the survey type

Section titled “3. Select the survey type”

Choose the appropriate type for your objective:

TypeUse when
Brand trackYou want competitive benchmarking across the brand funnel. Questions are locked and standardised.
SegmentationYou want to understand how a segment behaves inside a specific context (e.g., in-store behaviour at Walmart).
CustomYou have your own research questions. Build and order them freely.
Quick / PollYou need a rapid, simple preference question with a few options.

4. Pick your segment and build or confirm the question set

Section titled “4. Pick your segment and build or confirm the question set”

Select which segment’s respondents will take the survey. For standard surveys, confirm the question set. For custom surveys, write your questions.

When planning questions, think about what you are trying to learn from each segment. Two segments may get different questions even within the same study because their mindsets and relationship to the category differ.

5. Set the respondent count

Section titled “5. Set the respondent count”

Size the respondent pool appropriately for your population. For US studies, 10,000 respondents is generally sufficient. For smaller or more niche populations, fewer respondents may be appropriate; sample size depends on the population you are representing, not a fixed rule.

For a detailed explanation of sampling logic and how design effect and margin of error interact with respondent count, see Statistics.

6. Run the survey

Section titled “6. Run the survey”

Confirm and launch. The platform runs the survey against your respondent pool automatically.

Reading the results

Section titled “Reading the results”

When results are ready, see Read & Interpret Results for the full guide. At a high level:

  • Distribution view: bar charts showing how respondents answered each question.
  • Story / narrative layer: the platform identifies a leader and follower framing (e.g., “Apple leads by a wide margin; Samsung is the closest follower”) and surfaces the patterns in plain language.
  • Header stats: total respondents, universe size, margin of error, design effect.

What comes next

Section titled “What comes next”