Conjoint Analysis for Marketing Research

Navigation: Main page » Marketing

 Print this page Submit articleComment article

Increase Traffic Network - Get Targeted Traffic Right Now!
Need web traffic. Try "Increase Traffic Network" Today. Get the targeted web traffic you need. You get: * Real Time Statistics * 24 hr Unique Visitors * Traffic Targeting * Gauranteed or your Money Back Pick The Targeted Category. Pick How Much Traffic You Want! Submit Your Funds. Submit Your website data. Your Done! Traffic is on the way!

Author: Michael Petty

Conjoint analysis is one of the most popular quantitative methods in Marketing Research. The popularity of Conjoint Analysis is well deserved. Conjoint Analysis has in many cases replaced concept testing because it is more effective. The problems with traditional product testing were shown dramatically in 1985 when Coca Cola carried out market research at a cost of $4 million, looking for a new formulation to replace the classic Coke. Despite the fact that the New Coke was preferred to Classic Coke by 55% of people in taste tests, the alienation of a significant number of consumers who preferred the classic formulation resulted in a public outcry by Coca Cola loyalists. Basically they were saying to America “don’t mess with our Coke” . The major reason for the failure of the research into the New Coke was that it was assumed, incorrectly, that a better tasting formulation â€" more like Pepsi â€" would sell more Coke. This is a basic assumption that underlies most conventional concept and product testing. The faulty research failed to reveal that the strength of the Coke brand far outweighed small preferences in taste. In fact in carrying out typical blind product testing Coca Cola management had badly underestimated the power of their brand. As Donald R Keough, President of Coca Cola said in explaining the research debacle: “We did not understand the deep emotions of so many of our customers for Coca Cola”. Direct questioning like the following used in product testing leads to inflated estimates of importance. Usually respondents are asked:

In buying a small car how important is it to you that it gets high mileage per gallon?

They are then asked to indicate the importance on a five-point scale ranging from 1 = Not important to 5 = Extremely important

But if we also ask about the relative importance of styling and warranty then we shall likely find that these matters are just as important as m.p.g.

The conjoint task is more realistic. Respondents may be asked: Which of the following small cars would you choose to buy?

(01)Car

• 2009 Chevrolet Aveo5
• MSRP $11,695
• 1.6-liter, 107-hp four-cylinder engine
• 27 mpg city/34 mpg hwy
• AM/FM stereo with AUX jack
• 3 year/36K warranty
• Comfort and quality rating 8/10
• Transmission â€" manual

(02)Car

• 2009 Kia Rio5
• MSRP $13,325
• 1.6-liter, 110-hp four-cylinder engine
• 27 mpg city/32 mpg hwy
• AM/FM/CD/MP3/SIRIUS, USB-AUX jack
• 5 year/60K warranty
• Comfort and quality rating 7/10
• Transmission - manual

Conjoint analysis has developed in the past two decades to become much easier to use in research. Conjoint analysis has always been a powerful research technique because, rather than asking consumers to rate the attributes of products, something consumers are not particularly good at, it presents products as a bundle of attributes and asks respondents to choose between the products. Thus the conjoint task for consumers is very similar to what they do when shopping. Conjoint analysis uses powerful experimental designs and statistical techniques to measure the effect each attribute of a product has on consumer choices. Conjoint designs lead to much more powerful research than the typical ex post-facto designs of typical concept research. Conjoint analysis can:

• Optimize a product line
• Measure the strength of a brand and learn how exploit the brand’s strength
• Estimate price elasticities of demand; this allows a company to construct models of demand for products at different prices in existing markets or in new markets
• Segment customer bases to estimate, for instance, the size of a market segment made up of brand loyalists.

Michael Petty PhD www.lasir.info