|A new web-based conjoint product from US brand development specialists Moskowitz Jacobs, seems set to bring this under-used research technique out of the shadows by offering affordable DIY conjoint interviewing and trade-off analysis online.
Often used on brand image research, conjoint research has suffered from a bit of an image problem caused by perceived high cost, difficulty of execution and, perhaps most damaging, lack of understanding in the research community. IdeaMap.Net effectively solves the first two problems with a simple set of tools that let you drive the whole process yourself from a web browser on a modest pay-per-survey basis. It is also very fast: conjoint studies among thousands of respondents can be turned round in just 48 hours.
There are three main components to IdeaMap.Net: a design tool, the survey engine and a range of back-end analysis tools, giving it a superficial similarity to a hundred other web interviewing tools.
But there the similarity ends. Instead of defining questions, you define elements and categories (groups of related elements), which combine to create "concepts".
Conjoint research provides a more rigorous alternative to large swathes of conventional research often tackled by batteries of satisfaction and importance ratings. Elements from the different categories you create are combined before the respondent in a systematically varied way, and an affinity rating is collected for the entire "concept" - generally using a nine point scale. It aims to simulate more closely the multi-dimensional view the consumer always has of any product, brand or service. In IdeaMap.Net, four pre-defined models give different numbers of element/category groupings at four different price points.
A complex and proprietary statistical algorithm then carries out a trade-off analysis on each respondent's data and breaks this down into the individual preferences for each element. The results can be cross-tabbed against demographics in the conventional way. IdeaMap.Net provides a decent range of back-end analysis tools, and will also export data out in SPSS or SAS format.
|IdeaMap.Net: web-based conjoint product
|It is probably best to start designing your study in Word or Excel, as once you have chosen your number of elements it cannot be changed. The interface is very simple and obvious to any web-literate researcher. Element texts and images are easily pasted in or uploaded from an image library. Single-coded demographic questions are also easy to add, though the system would benefit from the ability to copy blocks of question or entire questionnaires from past projects. You are prompted to create a welcome screen and a concluding screen for your survey, and can set an option to close the survey by going to another web page.
Going live is also very simple: clicking an icon will activate the study and present you with its unique web address and password. This can be pasted into an invitation email, though you need to do the emailing yourself or get your sample supplier to do it for you. Security is rudimentary though probably sufficient: though there is only one password per survey, cookies can be enabled to prevent multiple participation.
However, once live, you cannot change as much as a single spelling mistake without having to re-program the entire study from scratch - it tends to make you rather diligent in your use of the excellent preview facilities the tool provides.
Ford & Earl is a design company based in Troy, Michigan, specialising in brand building. Consumer Insight Manager, Renee Cameron, has put hundreds of surveys through IdeaMap.Net.
"The appealing thing about this method is that, for the price of one focus group and in less time, you can have
|quantitative data. That's appealing from the cost and time standpoint. But even more appealing is the integrity of the methodology."
For Cameron, conjoint is better at "picking winners" than five point rating scales. "So often, if you look at the 'top two box' everything has the same mean. Everything scores 4.25. But if everything is rated the same, that is not an actionable finding."
IdeaMap.Net, she feels, provides two levels of learning from each study. "One is the clear winner for this study, this client, now; and a second level of analysis which we call the meta-analysis, where we convert, say, 40 datasets into one." From this, she can test other hypotheses that have started to emerge during the previous studies - without the need to gather more data.
"It's a method that is designed to be iterative, and dynamic so that the learning never stops," says Ms Cameron. "And it continues to amaze me that I can send out 8,000 email invitations at 4pm, and when I get back in the morning, my project is finished!"
Although performance like that does depend on having a target audience with good representative internet samples, there is nothing to prevent IdeaMap.Net working well in hall tests or kiosk settings, with a little imagination.
Tim Macer writes as an independent specialist and advisor in software for market research. His website is at www.meaning.uk.com