New Product Management (7)

Chapter 7: Analytical Attribute Approaches: Trade-off Analysis & Qualitative techniques

Trade-off analysis (often called conjoint analysis) – it is more common used in concept evaluation, used in generating high-potential concepts for future evaluation. Trade-off analysis refers to the analysis of the process by which customers compare and evaluate brands based on their attributes/ features. Conjoint analysis: one of the most common analytical tools used to assess tradeoffs (like factor analysis which is a tool used to develop perceptual maps – tradeoff analysis is the broader term). With conjoint analysis, we can represent a product as a set of bundle of attributes. It puts all of the determinant attributes together in new sets, and identifies which sets of attributes would be most liked/ preferred by customers.

Using trade-off analysis to generate concepts

Coffee has three determinant attributes: flavor, strength, and intensity of aroma. If we could get customer preferences (or utilities) for each attribute, we could combine the best level of each attribute into an overall favorite product. Unless this particular combination was already on the market, we would have our new product concept. Cards representing new products that combined specific features, functions, and benefits would be prepared, and customers were asked to sort the cards by preference from top to bottom. As business buyers tend to make a more rational analysis of product features, trade-off analysis is becoming increasingly valuable for industrial product innovation.

A conjoint analysis application

Supposed our understanding of the market and recent customers research, we have found that three attributes are uppermost in Customers’ minds when they buy Salsa – spiciness (mild, medium, extra hot), color (green or red), and thickness (regular, thick, extra thick). There are three x 2 x three = 18 types that can be made. Then 18 cards will be designed (each with pictures or verbal description), and each respondent will be asked to rank the cards from 1 to 18 where 1 is “like most”. The largest range in utilities is found for spiciness, thus spiciness is this individual’s most important attribute influencing likes and dislikes – while color is the least important. This analysis uses monotone analysis of variance (MONANOVA), a data analysis technique, to find these patterns within the rank order data. Trade-off analysis allows the manager to identify which attributes are the most important and which level of these attributes are the most popular. However, it may be about half the market likes the mild and the other half goes for extra hot, and if we examine the average and conclude that medium-hot is best, in reality nobody may like it. Moreover, if there are more attributes and levels – there could be more than 400 cards to rank. We must choose the combination that most closely meets their desires by trading off attributes wanted more against those wanted less.

A guideline of conjoint analysis includes:

1) Able to specify the product as a bundle of attributes (difficult for perfume as in factor analysis)

2) Know what the determinant attributes are before we do the conjoint analysis (AR gap mapping can be helpful).

3) Respondents should be familiar enough with the product category and attributes to be able to provide meaningful data (conjoint analysis will be less useful in case of new-to-the-world products).

4) The firm should be able to act on the results (to actually develop a product).

Discrete choices method is used when dealing with the complexities of many product situations. Conjoint results are a valid early indicator of ultimate product success, at least for product line extensions (conjoint analysis, perceptual mapping, as well as product trial will be rich sources of customer info later in the new product process). In the case of major innovations, customers without a high level of expertise may be unable to access the innovation’s benefits, and concept test results may not be validly predict how well the actual product will be received.

Virtual prototypes in concept testing

There are either static pictures of the prototypes or video clips that stimulate the product in action (less costly to produce and test than actual physical prototypes). Information acceleration (IA) – developed & 1st applied to GM, the respondents are brought into a virtual buying environment that simulates the info typically available in a realistic purchase situation. They see ads, read car magazines, and hear statements from salespeople and Word of Mouth. Customers could walk around the showroom and look at computer generated car prototypes.

Qualitative techniques

1) Dimensional analysis – listing all of the features of a product type. The product concept creativity is triggered as we instinctively think about how that feature could be changed. Just by citing a unique dimension could spark ideation.

2) Checklist – today’s most widely used idea-generating techniques – 8 powerful questions which lead to useful ideation: Can it be adapted, modified, reversed, combine with anything, magnified, minified, rearranged, and can something be substituted? It produces a multitude of potential new product concepts (though most are useless). However, one attribute could be made dependent on another (a child’s bath mat changes color if the water is too hot), an attribute might be removed, leading to an essentially different product.

3) Relationships analysis – 2 analytical attribute techniques: the 2-dimensional matrix, and the morphological matrix. Both require the respondent to find relationships among dimensions to generate new product concepts. Attributes are features, functions, and benefits – but other aspects of products are not always included as attributes in definitions (different places of use, occupations of users). Relationship analysis uses them too. We seek all dimensions that help, and there is no fixed set of these.

i) 2-dimensional matrix – relationship analysis goes direct to a new product idea. The number of these matrixes that can be prepared is almost unlimited. A slightly different kind of relationships analysis employs as its dimensions the product’s utility lever (how the product affects the customers’ life) and buyer’s experience cycle (at what point does the product affect the customer).

ii) Morphological/ multidimensional matrix – simultaneously combines more than 2 dimensions. The matrix can include many dimensions. Dimensions could be cleaning instrument (broom), ingredients (alcohol), objects to be cleaned A/C filter), package (aerosol), substance to be removed (blood), and texture (cream). What at 1st appears to be noise may simply be a great new idea no one would have thought of easily without the matrix.

4) Analogy – we can get a better idea of something by looking at it through something else. It is used heavily as part of the problem-solving step in problem-based methods. Preparing, serving, and consuming meals in a plane are clearly analogous to doing so in the home (result in the good ideas for new processes in the home kitchen). Amusement park designers watched cattle being herded and came up with queues for those waiting to go for a ride. An analogy for bicycles might be driving a car (incorporate steering, moving, slowing, curving and others – but autos have 4 wheels for stability, built-in communication). The analogy should meet 4 criteria: vivid & have a definite life of its own, full of concrete images, should be a happening – a process of change/ activity, and a well-know activity and easy to visualize and describe.