Analysis Techniques

The following are techniques we will likely use to accurately and insightfully analyze your data. Though these methods may be complex, the deliverable you get will be simple to understand and actionable.

Text Analytics

  • Used on qualitative data to understand what respondents say and how they say it
  • Measures the frequency with which respondents use certain words and concepts when discussing a topic
  • Determines the extent of respondents’ positive and negative emotions when talking about a topic or a company

Correspondence Analysis

  • Can identify attributes that are most uniquely associated with specific brands
  • Commonly applied in positioning research to create perceptual maps showing each brand’s position


  • Utilized to compare the differences in responses between groups of respondents
  • Helpful in reliably determining if certain segments are more likely than others to purchase an offering, prefer a brand, use a feature and so on

Regression Analysis

  • Implemented to examine the relationship between dependent and independent variables
  • Useful in pinpointing the specific features or attributes that drive overall satisfaction, loyalty and advocacy

Max-Diff Analysis

  • A simpler form of tradeoff analysis that has respondents indicate the importance of attributes by making choices between them
  • Particularly valuable when having respondents react to a long list of attributes

Conjoint Analysis

  • A more complex tradeoff analysis that is typically used to predict the desirability of combinations of attributes as would be found in a company offering
  • Valuable in identifying the best configuration for an offering, estimating how an offering might perform against competitors, and determining optimal pricing

Factor and Cluster Analysis

  • Examines the complex relationship between multiple variables, condenses them into smaller groups of factors, and assigns people to unique groups
  • Often employed in research to identify segments that can be targeted and addressed differently