Feature Value Modelling

Feature Value Modelling; The Added Value in the Consumer Durables Market

by Maureen Bannink and Jürgen Warnecke, SKIM Analytical - Durables & ICT

Knowledge and insight regarding the consumer perception of product features is vital in the consumer durable industry. When developing new products, the manufacturer has to decide which features should be implemented. When marketing (new) products, the marketer has to decide which product features should be brought to the attention of consumers and which should be ignored in the communications efforts.

In product development, the interests of the manufacturer and the consumer are opposite:
A manufacturer wants to reduce the production cost of products as much as possible (which usually implies equipping products only with basic features) and sell them at the highest possible price, while a consumer wants to buy products with the highest added value (usually this implies ‘nice' features and therefore high production costs) at the lowest possible price.

It is clear that both parties need to compromise. E.g.: it is very likely that a consumer is prepared to settle for a less preferred feature, if the product is improved in another aspect. To determine an ‘optimal compromise', the manufacturer needs insight into how a consumer decides on a particular product. In other words, get insight into the importance of features relative to each other, but also relative to price.

Methodology
The easiest way to establish the importance of features would be simply asking a representative sample which features they think are important and (by means of a scale) how important these features are. However, it can be expected that a consumer would regard all features important and the key question, how people compromise or make trade offs, will not be answered. We have therefore employed Feature Value Modelling (FVM).

The cornerstone of FVM is Choice-Based Conjoint (CBC). In CBC, we mimic reality as much as possible: we ask a respondent to imagine he is buying a product, for example, an MP3-player. We show him a number of players, defined by a combination of key features. These combinations of key features can include all product characteristics, product features as well as services.


Example of choice task as can be used in the field

This process should be repeated a number of times, therefore we offer the respondent a series of choice tasks. Each time, the respondent will optimise as much as possible by selecting the best MP3-player, based on his personal preferences. Afterwards, we analyse the choices made and thereby quantify, on an individual level, the preferences regarding the features as well as the relative weight each feature has when making a choice regarding an MP3-player. These quantified preferences and weights are expressed in utilities.

Though conjoint analysis is a very powerful technique, we have to note that it only measures the importance of features: it does not explain the importance of features. When it comes to consumer durables, most consumers decide on a ‘benefit level'. In other words, they have certain preferences regarding particular features because these features provide certain benefits (e.g. consumers prefer a smaller and lighter MP3-player because it can easier be carried along or because it is more modern or because it makes the product more aesthetically attractive, etc). To understand why people prefer certain features, we include a series of benefits (per feature) in the FVM questionnaire. Consequently, we can determine the differentiating benefits each feature brings.

Deliverables
A database full of utilities on an individual level would not appear very useful by itself, but when analysed it offers an abundance of useful insights:

First of all, you can determine how you can make your products more attractive and how each alteration by itself or combined will attribute to the perceived value of the products. This enables you to make the trade-offs necessary when developing a new product. Furthermore, brand equity can be measured, as well as price sensitivity (how effective will lowering the price of product X be).

Secondly, the results can be used for segmentation purposes; identifying segments based on similarities in preferences and derived from actual choice behaviour rather than demographics or stated preferences.



The ‘best' about Feature Value Modelling is the market simulator, which ‘runs' on the database of utilities and allows forecasting. The procedure is relatively easy. First, all relevant products in the simulator are defined by a combination of features. These can be your own products, as well as competitor products. After that, the simulator relates each product to the utility scores for each feature for every respondent. The scores per product are added and it is assumed that a consumer would choose / buy the product that has the highest total utility (the most value for money). The output of the simulator is the share of choice (percentages of consumers that would choose a certain product) for each of the products included in the simulator.

Existing products and product lines, as well as concept products can be tested in the simulator . Accordingly, the potential of a new product in the market can be established very easily. In addition, cannibalisation effects can be analysed, as separate simulations ‘before and after' the (simulated) introduction of a new product can be run. The simulator can also be used as a valuable tool when developing contingency strategies. You can, for example, test the impact of a price reduction by your competitor on the share of choice of your products. Moreover, you can test the effectiveness of alternative ‘counter strikes'.


Example of share of choice chart, based on FVM

Analysing the differentiating benefits enables us to truly understand why some features drive choice, whereas others have no impact at all. Additionally, we can give input regarding which claims and USPs to stress on the crucial benefit level when communicating a (new) product in the media.


Example of differentiating benefits chart, based on FVM

About SKIM Analytical
Since its foundation, SKIM Analytical has operated on the cutting edge of technology-driven and market-driven product development and marketing. Our Durables & ICT division specialises in providing strategic information on the consumers point of view regarding products and services in a pre-launch stadium. SKIM Analytical has vast experience in conducting multi-country research and we regard the application of advanced research methodologies and techniques our core capability.

With the backgrounds of our staff, ranging from corporate communication sciences, to business economics and business and industrial statistics, we can approach marketing from all relevant angles. SKIM Analytical Durables & ICT is a young, independent and flexible research team that is truly dedicated to delivering the necessary consumer insight to our clients.

For more information about Feature Value Modelling, please contact:

Maureen Bannink
Division Manager
SKIM Analytical - Durables & ICT
+31 (0)10 282 3535
durables&ict@skimgroup.com