A breath of fresh air

As the world’s largest maker of air conditioning, heating and refrigeration equipment, Carrier Corporation is no stranger to global commerce. The Farmington, Conn.-based company does business in 167 countries. Its products are designed and manufactured in sites around the world.

Carrier had conducted trade-off research on its line of hi-wall ductfree split air conditioning units, which are used in commercial and residential settings, but that research had focused on issues like efficiency, sound level, size, etc. As designers readied new models, the company wanted to find out which control features (things like timing functions and air distribution options) would be most desirable and feasible, given the amount of money budgeted for control features during product development. In addition, was there an opportunity to make Carrier’s products stand out in the individual foreign markets by offering features customized to those markets?

"Our products and those of our competitors offer a menu of features. We wanted to look at those features currently offered, as well as others we thought were possible, and try to understand which ones were most valued by customers," says Bob Whitwell, director, global ductfree split systems, Carrier Corp.

"We wanted to see if we could use the same product in every market and, if so, what the potential impact of doing that would be on customer satisfaction. We also wanted to understand if there were opportunities for differentiation in individual markets by providing some features that aren’t generally valued on a global basis."

Kano method

Rather than a standard conjoint study, which it had used in the past, Carrier and its research partner, BAIGlobal, Tarrytown, N.Y., conducted face-to-face interviews with potential commercial and residential users of the ductfree systems in Italy, Spain, Korea and Singapore using the Kano method.

Based on the work of Noriaki Kano, a professor at Tokyo Rika University, the Kano method aims to uncover the subtle reasons why certain product attributes are more desirable than others by requiring respondents to answer two-part questions for each attribute: "How do you feel if a feature is present in the product?" and "How do you feel if the feature isn’t present?" Respondents must answer each part with one of five replies:

  • I like it that way.
  • It must be in the product.
  • It does not make a difference to me.
  • I can live with it that way.
  • I dislike it that way.

The Kano method tells you how attractive a feature is and how a person views it by classifying it into one of six categories based on the combination of answers to the two-part questions:

  • Attractive - The customer is more satisfied when the product has this feature, but is not necessarily dissatisfied if it doesn’t. On a car, for example, an automatically retracting radio antenna is nice to have but its absence wouldn’t make a person choose not to buy the car.

  • Must-be - The product must have this feature or the consumer would be dissatisfied, but the consumer is neutral about it otherwise, because it’s an expected feature. Continuing with the car example, you expect a car to have good brakes.

  • One-dimensional - The more of the feature, the better. The better a car’s gas mileage, the happier the consumer is. If it doesn’t have it, people are dissatisfied.

  • Reverse - The customer does not want the feature and having it means dissatisfaction. An example might be a car’s color.

  • Indifferent - The customer doesn’t care either way. Having the feature doesn’t mean satisfaction or dissatisfaction.

  • Questionable - The customer’s responses on this feature contradict themselves.

"We had proposed using conjoint but there was a curiosity about this methodology," says Gunilla Broadbent, president of the Worldwide Services Division of BAIGlobal. "The client wanted to see if it might give more nuance. Conjoint gives you values which are linear, which is an advantage. You can really look at the difference in importance between attributes because the level of the utility is totally linear, which makes it easier to interpret but isn’t always totally correct, because there are certain things that, from an emotional point of view, might be more or less important to a person but they don’t come out."

"We were trying to look for an alternative that wouldn’t involve the cost and limitations of the traditional types of trade-off, conjoint and discrete choice," says Joe Lanzetta, director of global marketing research, Carrier Corp. "This approach allowed us to construct a survey where respondents could answer questions on 15 or 20 attributes and give us an idea of the importance of each. In the past we had used conjoint and discrete choice, which are good techniques, but one of the limitations is the number of attributes you can efficiently test. You can test a lot of attributes but it requires setting up more and more scenarios, and that can be time-consuming and expensive."

Broadbent says that while the Kano method does provide insights to the subtle reasons why one feature is preferred over another, the technique isn’t a replacement for conjoint. "This method presents a challenge, especially when you’re dealing with several different languages, to find the right nuance and the right way of expressing it, so we did a lot of pretesting before we were satisfied with the wording. It’s very difficult to always find exact equivalents to what you had in English, and here it’s especially important because if the nuance is wrong, the whole analysis becomes wrong. That’s a drawback to the technique; it is very sensitive to language."

"We had to be a little more careful on translation because we wanted to maintain the meaning of statements like ‘I like it that way.’ You had to make sure that you captured the essence," Lanzetta says.

Measuring price sensitivity was another facet of the research. Respondents were asked to indicate a range of prices, including prices at which they would view the product as so inexpensive that it might be shoddy, those at which it’s a bargain but a quality product, and those at which it’s unaffordable. "As the manufacturer, you have to define the prices, which may be difficult. This way the consumers define the price for you. By asking people how much they’re willing to pay you probably get an unrealistic price, but with the approach we used, you understand where the person is coming from, what is too cheap or too expensive, how big the normal price range is for that person and where it falls," Broadbent says.

Biggest lesson

Working on this study, his second international research project at Carrier, Whitwell says he learned a lot. The biggest lesson is that it’s helpful to have a research company handle the overseas interviewing and coordination. "For the first international project, we took the coordinator role and contacted individual research companies in each of the countries to do the field work. It was not an easy thing for us to do. We ended up spending a lot of time analyzing the results across countries and trying to draw conclusions on a global basis. In this case we used BAIGlobal to perform that function. That made it simpler for us. They had companies that they were familiar in each of the countries that we wanted to conduct research in so they handled a lot of those issues."

Client and research company worked to make sure the translation was accurate. After the surveys were translated into the local language, representatives from local Carrier offices in those countries checked the surveys to make sure the technical terminology was translated correctly.

In addition, usage examples and scenarios were tailored to the individual markets, to make them relevant, Whitwell says. "We wanted to give them a good baseline from which to make their comparison. For example, when we talked to commercial customers, we gave them a price and a system capacity that they would commonly consider putting into their application. We did the same for the residential customers. In Italy, for example, the typical application for a residential customer might be a bedroom, so he’s going to have a different capacity size than the residential customer in Korea, who might place it in his living room, and the pricing would be different, so we came up with a range of prices and applications depending on who was being asked the questions."

Opportunity exists

The research gave Carrier a good handle on the importance of the various features and confirmed company suspicions that it might be worthwhile to begin marketing products with features customized to the individual markets and applications, Whitwell says. "We had budgeted a certain amount of money in the cost of the product for the control features, and we wanted to provide the most valued features that we could for the cost that we had budgeted. If it was a ‘must-be,’ then we had to figure out the most cost-effective way to provide this feature. If it’s ‘one-dimensional,’ we have to make a judgement as to how much we can spend on it and what level to provide. ‘Attractive’ features we could put in if we had money left in our controls budget; ‘neutral’ we wouldn’t spend money on.

"There was a preliminary list of features that we intended to put in the product. In this case the product development cycle was ahead of the research so we said let’s use this tentative list of features and start the development of the product and use the research to validate. If we find something different as a result of the research we want to be prepared to go back and make the appropriate changes. For the most part, our initial choices were validated."