Whether your research requires a straightforward cross-tab (univariate) analysis of total respondents and key subgroups or the consideration of more insightful multivariate analysis techniques, Vault will work with you to develop a program to best meet your needs.
We have extensive experience with multivariate data analysis/techniques, including:
- Ideal Analysis
- Concept, Product, Advertising Optimization
- Perceptual Mapping
- Conjoint/Discrete Choice Analysis
Concept, Product, Advertising Optimization
- A Key Driver Analysis will further clarify key components that drive interest/motivation – overall and potentially among any particular segment. The technique uses regression analysis to provide a model that identifies the weight, or impact (derived importance), of any given attribute on overall concept (or product) interest. When overlaid against stated importance, we will be able to identify the expectations that may be unable to be articulated, yet recognizable and appreciated once identified. The analysis results in the identification of attributes that are perceived to be essential, cost-of-entry or implied vs. those that they are indifferent to.
- A Quadrant Analysis provides a two-dimensional representation of importance vs. performance ratings with regard to various attributes. Respondents are asked to rate a series of attributes with regard to the importance they place on that attribute, as well as how they rate a brand on that same attribute. This is one of several ways to approach “brand vulnerability” from an analysis standpoint. Quadrant analysis typically results in the identification of one or more attributes where a brand outperforms alternatives/competitors on attributes considered important. Conversely, under performance on key attributes can also be identified with an eye toward improving perceptions/ performance. These valuable insights will enable [brand] to:
- Effectively address concerns of current customers
- Convert lower volume / non-engaged customers into higher volume / engaged customers
Perceptual maps show the position of a given product/brand relative to others currently on the market. Additionally, a perceptual map will show what features/changes/additions could “move” the brand in a specific direction. This results in the assessment of a brand’s perceived strengths and weaknesses against those of alternative options/competitors.
Often used for prioritizing a large number of attributes/benefits/brands, the max-diff exercise is based on a measure of customer choice and trade-off, rather than standard rating scale responses. Respondents evaluate multiple sets of 4-6 attributes/benefits/brands, indicating both the most important and the least important. Responses are analyzed using various regression model techniques to derive attribute importance scores at the individual respondent level. The results are used to predict future customer behavior. This technique allows a researcher to test a large number of choices relatively quickly.
Conjoint/Discrete Choice Analysis
Discrete Choice Analysis is a type of Conjoint Analysis designed to model buyer choice. Through the use of choice data, we can determine which features of a product drive choice. In this method, respondents are given the ability to choose between distinct (“discrete”) sets of options that closely parallel the true decision-making process that buyers are faced with. As such, the options assume levels of mutual exclusivity. The power of the technique is derived from these distinct options.
- To buy brand A, brand B, some other brand, no brand.
- To travel by car, train, plane, bus.
- To vacation in Europe, Asia, elsewhere, or to stay home.
- Note that not all choices are discrete, e.g., filling a gas tank or some lesser amount.
The data output is an optimal product profile(s). The data, in some circumstances, can also be used in forecasting simulations.
There are a wide range of definitions and analysis options when it comes to “segmentation.” These options can be grouped according to two basic types of segmentation:
- A priori — defined as formed or conceived beforehand
- Examples of a priori segmentation schemes include:
- Usage (e.g. heavy vs. light users)
- Demographics (e.g. older vs. younger)
- Customers vs. non-customers
- Intent to purchase
- Post hoc — defined as formulated after the factExamples of a post hoc segmentation schemes include:
- Cluster analysis
- Latent class analysis (LCA)
- Discriminant Analysis
- Classification Tree Methods — e.g. CHAID, CART
- Factor Analysis
Regardless of approach, some potential benefits of segmentation include:
- Identification of:
- Highest value customers to target
- Marketing niches, enabling targeting of customers with less competition
- Ability to (by segment):
- Target communications
- Target promotions
- Develop different products