Analytical Toolbox

CRS employs leading-edge analytical techniques to enhance the value and actionability of the research insights.  Over 80% of our projects employ some advanced statistical analytics to help uncover the underlying reasons or relationships in the data.  We always try to de-mystify the techniques so they are as user friendly as possible.

It is our belief that there are a variety of different analytical tools that can be used to help answer the research objectives.  Some statistical techniques group people, some group attributes, some establish priorities or degrees of impact, and some force trade-offs.

CRS selects the best approach for each individual situation.  Sometimes this requires running the data using several different statistical techniques and finding the one that produces the best fit and gives the most intuitive and actionable solution.

We like to think about analytical techniques in terms of applications and the problems they help to solve.  For this reason, we group our statistical tools and techniques into six basic categories which are shown below:

Attribute Prioritization Techniques:

These tools are used to help categorize, identify or prioritize a battery of attributes.  The attributes might be product/service characteristics or attitudes and lifestyles.  In many cases, the results are presented in simple quadrant charts.  The types of attribute prioritization techniques used at CRS include:

  • Discriminating Quadrant Analysis (DQA)
  • Unarticulated Needs Quadrant Map
  • Opportunity Priority Quadrant Analysis (OPQA)
  • Product Modification Quadrant Analysis
  • Maximum Difference Scaling (MaxDiff)
  • Penalty Analysis

Trade-Off Methods:

This family of techniques is used to help understand the importance or contribution a feature or attribute makes on a product/service.  Understanding the “utility” of the feature allows for trade-offs or the optimizing of the feature set.  The relative predicted take-rates for the created feature sets are an output of nearly all the techniques CRS employs.  We also use these trade-off methods to predict demand elasticity and price optimization.  The types of trade-off methods used at CRS are as follows:

  • Conjoint Analysis
  • Discrete Choice Modeling
  • Choice-Based Conjoint Analysis
  • Constant Sum
  • Maximum Difference Scaling (MaxDiff)

Segmentation Methods:

Segmentation is used to identify groups of people who have similar characteristics.  These segments can then be better targeted with products/services and marketing communications.  Our segmentation studies often include segmenting on two or three different basis variable sets and then cross-tabulating each of the segmentation schemes against one another to develop rich segment profiles.  We regularly segment on needs, attitudes, deficiency gaps, occasions, psychographics, utilities, or demographics.  Shown below are some of the segmentation methods employed.

  • Cluster Analysis
  • Latent Class Analysis

Optimization Methods:

These techniques are used to optimize specific product characteristics or for the optimization of product lines.  These analytical techniques are non-trade-off based approaches. The first two techniques are generally used for sensory related projects, while TURF Analysis is used in a variety of situations and types of projects.

  • Product Modification Mapping
  • Response Surface Methods
  • TURF (Total Unduplicated Reach & Frequency) Analysis

Demand Elasticity & Price Optimization:

These tools can be grouped into direct and indirect pricing methods.  They vary in sophistication, complexity and cost to administer.  CRS has had success employing all methods and selects the most appropriate technique to fit the study objectives and budget.

  • Direct Response Price Elasticity Modeling
  • Van Westendorp Price Sensitivity Model
  • Discrete Choice Modeling
  • Choice-Based Conjoint Analysis
  • Direct Response/CBC Hybrid

Positioning & Similarities:

Most of these techniques are based on competitive performance evaluations on a series of attributes which results in the inference of brand similarities and positioning.  Others are based on measures of direct brand similarities comparisons.  All of the techniques result in some type of perceptual map or quadrant chart.

  • Positioning Gaps Quadrant Chart
  • Discriminant Based Maps
  • MD-Pref Maps
  • Correspondence Based Maps
  • KYST Similarities Maps/Profit

24644925_s

 

Impact of Deals

 

 

26329792_s

 

 

Hierarchy of Traits

 

21359141_s

 

Price Elasticity

 

29612792_s (1)

 

 

Radial map