Methods of Driver Analysis
There are many different methods you can potentially use to undertake a quantitative driver analysis, and they fall broadly under two categories – we call them explicit and implicit.
Explicit methods involve directly asking respondents how important various attributes or conditions (drivers) are when choosing between brands or options.
An example is a stated importance question expressed as something like, “How important are each of these to you personally when considering product X?” and collected with a rating scale such as a 10-point scale (1 meaning “not at all important” to 10 meaning “extremely important”). These are simple to collect and analyse using nets – e.g. top 2 box (9-10) or top 3 box (8-10) – means and medians.
The stated reduction measure improves on this by asking respondents to select everything that is important from a multi-choice list, then presenting those choices in the next question which asks the respondent to select the most important. In this way, you get three tiers of drivers (“most important”, “important” and “not important”).
The issue with both of these methods is that there is a tendency for many things to be selected as important, making it difficult to know where to put your precious marketing and product spend.
Ranked importance is another popular explicit method. It involves asking respondents to simply rank drivers from most to least important. Although rankings can be useful to create some sort of hierarchy of drivers, they are flawed in terms of magnitude – we don't know how much more important each driver is from the next.
While appearing valid and useful at face value, explicit methods of driver analysis on their own are generally advised against by professional researchers. Enter the implicit methods…
Implicit methods include questioning techniques that effectively tap into and bring to light the more automatic or “heuristic” and passive ways that we make decisions about things, sometimes referred to as System 1 thinking (see Daniel Kahneman’s Thinking, Fast and Slow).
We have our favourite and not-so-favourite brands which can be based on things other than rational choice – such as how the brands make us feel about ourselves (image) when we use or are seen using them. Thus, to really get to the bottom of how choosers choose, we must also find ways to measure the more implicit or less spoken motivators.
Derived importance links the appeal or propensity of brands (likelihood to consider or purchase) with features or attributes associated with that brand. Put another way, this is the associated level of appeal for brands with a specific attribute – thus a derived metric. This can be done for multiple features using a range of brands in a category, and can be compared to a stated score to understand whether an attribute is overstated or understated, revealing even more about the type of buyer/chooser, and the realm of driver classification rather than hierarchy.
Regression analysis is a great modelling tool traditionally used for prediction by measuring the relationship between a set of independent variables (drivers or influencers) on a dependent variable (the thing you are trying to understand). It can also do driver analysis because those things which have a predictive value on an outcome can be viewed as drivers of that outcome. We can use this method to, for example, understand the drivers of Net Promoter Scores (NPS) – a metric which is simply a likelihood to recommend a make of car that you have owned to someone else – effectively a proxy for propensity. The downside of regression is that this is all predicated on regression finding a model that can explain variations in the dependent variable to an acceptable degree, and there is usually a need to experiment with different inputs and model variations.
MaxDiff (maximum difference scaling) is another popular methods that creates a hierarchy of preferences relating to a specific choice. It does this by forcing respondents to pick only the most important and least important item from a cycled list of attributes, helping to identify which they truly value. It can accommodate large numbers of drivers whilst still delivering rankings showing the relative importance of the items being rated. We love this approach and highly recommend it as a very reliable and valid driver analysis approach.
If you'd like a more comprehensive look at the different methods of driver analysis, including their pros and cons, download our FREE printable Driver Analysis Guide using the form at the bottom of this page.
Hybrid methods use a combination of stated (explicit) and derived (implicit) approaches to classify drivers rather than create hierarchies. (As such, we refer to them as "Im-Ex".)
A classification of drivers is, in many cases, actually more useful than a simple hierarchy because it potentially reveals how you should treat each driver and provides a lot more room to play with your product claims and marketing communications.
Im-Ex Quadrant is the simplest of the hybrid approaches and is loosely related to the Kano model of understanding customer preferences. It uses a simple stated importance measure but then adds a derived measure of importance and crosses them to create four quadrants of drivers. The downside of this method is that it tends not to work as well in markets where customers/users do not have a good knowledge and/or opinion of the competitive set. That said, this is still a very useful method, particularly in markets where there is strong competition and customers/users are likely to have formed opinions and/or have experience with a number of competitors.
Im-Ex Polygraph uses similar data as the Quadrant method but adds the concept of “attractors” and “detractors” to create a more sophisticated method of classification. The Polygraph essentially compares what people say they are attracted to or detracted by in a brand/option (stated importance), to the features of brands/options they are actually attracted to or detracted by (derived importance). In doing so, the analysis takes into consideration both cognitive biases and less spoken motivators.
As it turns out, the Polygraph in practice is a research output with a lot of utility because it enables a brand to focus on the core attractors and “wow” factors it can deliver, while worrying less about other areas that it might not have, but which it now knows fall in the “nice-to-haves” bucket. Conversely, it can avoid the areas that really detract (core detractors and “whoa” factors) while worrying less about some negative areas which, although detractors, turn out to be not as bad as stated. We have found this to be true across numerous categories and among brand custodians such as CEOs, CMOs, COOs as well as insights professionals.
Which method is best for you?
As we have seen, there are many ways and methods to do driver analysis from the most simple stated importance exercise where we get in-category respondents to rate the importance of certain attributes using a simple 10-point scale, to the more complex such as choice-based methods using maximum difference scaling and, of course, derived methods which link product attributes to a dependent variable such as purchase propensity or product satisfaction.
The chosen method will depend on many things and, of course, budget and timing constraints. In a perfect world, we would always move beyond just stated importance or rankings to ensure that we are tapping into consumers’ minds beyond just the rational thinking (System 2) into something that better approximates how choosers really do choose in the real world, incorporating our emotional and cognitive biases and use of shortcuts and heuristics (System 1 thinking).
Let D&M Research create and tailor a driver analysis study to suit you.
For a detailed guide to the methods used in driver analysis, and the pros and cons of each, download our printable Driver Analysis Guide
By submitting your details, you agree to receive further communications from D&M Research. You can change your settings at any time using the links at the bottom of our emails. We promise we won't spam!