By Gerard Loosschilder and Jemma Lampkin

In today’s medical device world, competition is fierce, and the pressure to innovate continues to increase. Development teams have to develop products that offer value to customers. More and more, the focus of innovation is at the feature level: designing winning or “killer” features that will entice customers to purchase the device.

Because innovative features are often associated with higher manufacturing costs, they must provide enough value to customers to justify a price premium. Beyond that, they need to be complemented with the right mix of product features.

There are many market research methods available to ascertain customers’ perceived value of specific features. One such method is Kano model analysis. This type of analysis is often used to support feature-level decisions: whether to include single features in a new product.

However, Kano model analysis can also help product management identify the optimal mix of features. Through Kano, an optimization procedure can be implemented to identify the combination of features that will generate the most excitement among customers and achieve the highest penetration into the market.

Classic Kano Analysis

Kano model analysis was introduced in the 1980s to set priorities among product specifications with the goal of launching a product that resonates most with customer preferences. The method is used to evaluate the aspects of a product one by one and to set priorities among the potential product specifications.

Kano model analysis can help product development teams identify not only the “must have” and “must not have” features of a product, but also the exciting features that help to differentiate the product competitively and enable it to command a premium price. This type of analysis can help device makers design a product with the right features, particularly ones that will excite customers, give the product an edge over the competition, and ultimately convince customers to purchase it.

Kano model analysis provides a framework to put features into the following categories:

Each feature is evaluated in terms of its potential contribution to future product success. When designing a product, the goal is to have a mix of exciting and must-have features while avoiding features that cause indifference or drive prospects away. This requires a heuristic from which to base decision-making about which features to include and keep out. For instance, one may decide to include all features that a sufficient number of potential customers consider a delighter or a must-have, while excluding all the features that a set number of potential customers call a must-not-have. The definition of a sufficient or a set number is a management decision, to be taken in the light of the results of the study.

To classify features, show customers a series of features, one by one, and ask them to answer a functional and dysfunctional question for each one:

For both questions, respondents should be asked to select one of the following answers:

By asking the two questions for each feature, a code frame is used to map the list of features into buckets and categorize them into one of the categories described above.

To quantify how strongly customers feel about their choices, Kano model analysis can be paired with traditional rating or ranking exercises. More recently, Kano model analysis has been combined with choice-based conjoint analysis, a statistical technique used to determine how people value a product’s or service’s features by determining the influence of those features on the choices of the customer. Regardless of which approach is used, the aim is to identify which potential product features should be included or excluded.

Using Kano for Product Optimizations

It would seem that Kano model analysis is most suitable to support individual feature decisions. However, this type of analysis can be taken a step further and also be used to identify the optimal product configuration. To do this, the results of a study using Kano model analysis are used to run product optimization exercises.

Optimizations are important for two reasons: to find the best product configurations and to prevent product developers from making mistakes. Say two features, A and B are each “must haves” to 50% of customers. By including feature 1, we may reach up to 50% of our customers. Similarly, if we include feature 2, we may also reach up to 50% of our customers. What happens if we include feature 1 and 2? Do we still reach 50% of our customers, or do we reach 100% of our customers (50% + 50%)? It depends, and the true reach is somewhere between 50% and 100%, depending how much the must-have preferences overlap between features A and B. 50% reach indicates a full overlap between features A and B, and a 100% reach indicates that the two features attract completely different customers. An optimization exercise takes the overlaps in feature preferences into account when identifying the feature combination with the higher reach in the customer base.

Defining the overlap is easy for combinations of two features. However, if the number of features added to the mix increases, the search for optimal combinations becomes tedious, and a more efficient procedure is needed. This is where running feature-based optimization comes in. A feature-based optimization procedure based on Kano model analysis considers two metrics for the features and their combinations, reach and excitement:

How It Works

Assume the medical device you are working on has five features to choose from. Any of them can be in or out, so the total number of possible products is 25= 32. The more features there are, the larger the number of possible products. The optimization exercise runs every possible combination of features. From this exercise, the top five combinations are identified in Table 1.

The “reach” row shows how many respondents in our study accept the product configuration. The rest do not accept it, choosing instead to either stay with their current product or buy a competing product. Reach can range from 0% (nobody) to 100% (everybody). Excitement also ranges from 0% to 100%, but the maximum excitement level of a product configuration is the same as its reach. So, if reach is 50%, excitement must be between 0% and 50%. This shows how excited the reached customers are. An excitement value of 33% at a reach of 53% is modest; it is not at its full potential (closer to 53%) but also not bad (closer to 0%).

The optimization example above shows that not all features need to be included in the final product. In fact, only feature 1 consistently appears among the top five features. Because it appears so consistently, it should be considered an exciter and a must-have feature in the product configuration. Conversely, feature 2 consistently fails to appear among the top five configurations. This feature not only fails to contribute, it seems to alienate customers. It is a dissatisfier and should be be left out.

Features 3, 4, and 5 appear in some of the top five solutions but not consistently. This means they do not contribute in a meaningful way to the product, rendering them indifferent. Not including them does no harm. Feature 4 is not even included in the best solution (#1). So, to conclude, the only feature that should be kept in the final product is feature 1, and the only feature that should definitely be left out is feature 2. The rest can be considered indifferent.

The Business Case for Better Feature Optimization

Because of the economic climate and new policies like the Affordable Care Act, attention is being focused not only clinical outcomes but also health-economic outcomes. Within the development of new medical devices and equipment, there is trend toward using value-based pricing, emphasizing the relationship between the cost of a medical device and its impact on the efficiency of care and return on investment. OEMs can no longer simply make claims about the clinical advantages of a feature or device; they must also demonstrate the impact of using them on the practitioner’s workflow, the quality and efficiency of care, patient and practitioner satisfaction, and ultimately, health-economic outcomes.

This sets the stage for running a feature-based product optimization. In the past, medical device manufacturers were often inclined to include every possible feature as a hygiene factor. Now that it has become impossible, we see a trend for market research to also support pruning decisions, helping management focus on features that are recognized to offer real value to all stakeholders, including healthcare providers, practitioners, and patients.

These decisions are not easy to make. In the face of increasing pressure to deliver health-economic value, product development teams can use feature-based optimizations to include features that excite stakeholders and that are recognized to deliver this value. However, it is also important to avoid pruning too much and risk alienating stakeholders by excluding features that matter to them. By using feature optimization to maximize value perception and minimize stakeholder alienation, medical device manufacturers are better equipped to make positive business cases for their products.

Gerard Loosschilder, PhD, is chief methodology officer and partner at SKIM, where he is responsible for innovation in methodology. Looschilder has more than 20 years of experience in market research in various positions in academia and on the client and the agency side. He earned a PhD in market research from Delft University of Technology in Delft, the Netherlands. Before joining SKIM, he was senior director of market intelligence at Philips domestic appliances and personal care. Reach him at [email protected].

Jemma Lampkin is a senior research consultant at SKIM. She has more than eight years of experience in the research industry, specializing in the healthcare sector the medical device market specifically. She has extensive experience designing and conducting global quantitative and qualitative market research studies in a wide range of healthcare indication areas. She holds a bachelor’s degree in psychology from Columbia University. Reach her at [email protected].

[image courtesy of CRAIGWBROWN/WIKIPEDIA]