Product Development Insight
Medical device developers need methods that enable them to get the most out of early-stage research and that will lead to strong, actionable data. Ideally, such data would enable designers to develop winning product designs that increase profits and meet caregiver and patient needs. One early-stage method is a technique that involves quantifying the voice of the customer. It is an expanded version of what is known as the semantic differential technique, and it can be used to capture the meaning of a given stimulus for users.
Such techniques are not new, but they often fail to get answers to the most critical questions required for understanding what the market wants and for understanding the users' needs and desires. The expanded technique provides a consistent system for eliciting quantifiable user input regarding aesthetics, emotions, attributes, user experience, and other aspects of the user experience before and during product development.
To use such voice-of-the-customer techniques, it is essential to understand how people rate products. Many years of empirical research have shown that people rate products along three key dimensions: power, activity, and evaluation.1,2 Such data can be measured three-dimensionally and can quantify the difference between a user's feelings about an existing product and a user's feelings about an ideal product.3 This information can identify opportunities for designers.
A Better Baseline
A measurement method must avoid placing the burden of design on the users and must not expect them to accurately describe features and aesthetics that they want in a product. The method discussed here allows users to respond to stimuli, which then defines how they react to a product. This information provides medical device designers with numerical data that can guide design.
Table I. This example of semantic differential scales shows labels that can be used at each point in the scale for a given product.
(click image to enlarge)
At its core, this technique captures the meaning of a stimulus to the user. It's important to note that there are not positive or negative ends to the scales used (see the example in Table I). For example, when evaluating an existing product (or a desired product) on a big-to-small scale, big isn't necessarily better than small; rather, the scale simply reflects how a person reacts to the product. The selection of just any scale won't lead to interpretable results. It's important to select scales that reflect the three key dimensions that people use to evaluate products.
Understanding the three dimensions for looking at products—evaluation, power, and activity—enables a medical device designer to create survey instruments that capture users' response to a stimulus (i.e., what a stimulus means to that user). The use of the term stimulus is intentional because this technique records the reaction to something rather than having a person answer a difficult open-ended question such as, “What product do you want?”
One key point is that if users are evaluating an actual stimulus (something that they can interact with), then they are evaluating their perceptions of the current stimuli. However, if users are evaluating an idea or concept, then they are indicating what their expectations are for that stimulus. This distinction allows for critical design insights. An example of this would be comparing the ideal ratings of nurses and doctors to identify ways to distinguish a new product from the competition. This method can also be used to compare environments to understand how products need to change to move from a medical environment, such as a physician's office, to home use.
Establishing User Expectations
One critical component in design is to establish users' expectations. Often designers develop what they think users will like, or worse, they design according to what they want (or hope) to make users feel—without understanding what experience those users are looking for. It is well known that users are very bad at articulating what they want, especially if they have little experience with a new product.
The semantic differential technique allows designers to have users describe the traits that define their ideal product. Defining user expectations does not mean measuring only their minimally acceptable requirements for a product. The scales compare user expectations for the perfect product against currently existing products (i.e., the manufacturer's current product and the competitor's product). The combination of scales indicates how far users' perceptions are from their ideal. The gaps in the ratings yield important insight into how a product can be designed or improved.
The data collected can be graphed in three dimensions (the evaluation, power, and activity dimensions are independent of one another) to illustrate how different products compare. Designers can also extrapolate the average distance from the ideal to evaluate the score repeatedly as a design progresses. Subsequent measurements show whether design iterations are progressing toward the goal (generally the ideal).
This technique is also useful for evaluating environments. Anything a user can have a reaction to can be evaluated. For example, to ensure that a medical device that will be used in the home actually fits in that environment, users can rate their home environment on the same set of scales (for current and ideal) to better understand what their expectations are for that environment. This gives the design team guidance for creating a design that fits with the users' expectations and goals.
Another use for this technique is to compare different user groups on their desired reactions to a product or environment. A recent project studied the ratings of family practice doctors (the likely purchaser of the product), as well as the nurses who would actually use the device throughout their day to understand whether their expectations for the ideal product were similar and, if not, where the differences may have existed. Nurses had higher expectations about the need for clear indications of the results (trust). Also, with more products going into the home health environment, it is important to consider how the home users (patients instead of healthcare professionals) want to react to a product in this environment. A great product design for a doctor's office will not necessarily translate well to a patient's home.
Identify the Gaps
It is essential to conduct this rating exercise before product development starts. The rating scales are used to elicit user reactions to a product, not to identify specific features. The technique is simple to implement.
First, it is important to understand the relative weight of each factor for the scales (evaluation, power, and activity). If all of the scales were to weight primarily on power, and none elicited feedback on activity, the resulting product might be superb in terms of power but fail overall in the market because the other areas were not considered. Therefore, it is essential to use rating scales that produce meaningful results by selecting scales that equally represent all three areas.4
The next step—collecting evaluations of competitive, ideal, and stereotypical products within a product category—highlights opportunities for innovation and determines the direction a new or existing product should take. In this step, users rate current, ideal, and competitor products. The distance (based on an average of the scales used to measure evaluation, power, and activity) between the ideal product and the stereotypical product presents the areas for development opportunity. If there is a large distance between those two points, it means that users are unfulfilled and expect more from such products.
However, if the distance between the ideal and the stereotypical products is minimal, it indicates that the product is likely a commodity and that there is little room for design improvement without making a giant leap in technology, features, or use paradigm.
The distance between a company's product and the stereotype is the novelty of the product, and the distance between a company's product and the ideal is the perceived room for improvement. Because dimensions are made up of specific rating scales, they allow the development team to hone in on specific areas for improvement.
Figure 1. Data plotted for the attributes of a new product along with a stereotype and an ideal shows the gaps between the attributes of the three designs.
(click image to enlarge)
Figure 1 plots the scale data for each product rated in a recent study. Notice the gaps between the new product and the stereotype. The distance from the stereotype is very favorable since the new product was rated closer to the ideal product, but there are still areas for improvement. In several cases, the new product was rated higher than the ideal, which suggests that the new one had too much of a particular characteristic (not that the new product was better). As the product development process continues, the new product is rated on these scales again to ensure that the design is moving closer to the ideal and closing the gaps.
Nurses: Finding the Right Warning
The semantic differential scales offer a great deal of flexibility. For example, one instance used this technique to have nurses evaluate different warning sounds. The goal was to choose the best warning sound from two alternatives. The study was to determine the perceptions and expectations of the warning sounds and how each of the two alternatives compared with those perceptions and expectations.
The study evaluated two different classes of warning sounds: a classic pulse and a melodic warning tone. Each class was composed of three alarms that represented low-, medium-, and high-criticality alarms (as defined by the duration and timing of the pulses). When designing a warning sound, designers can infer how they might want it to be perceived using the three key dimensions:
• Evaluation—the sound would be slightly bad. It would not be overly annoying, but not pleasant either, because then it would lose the desired connotation.
• Power—the power would increase (get stronger) as the criticality increased.
• Activity—the activity dimension would increase (get faster) as the criticality increased.
Figure 2. The scores for the low, medium, and high criticality for melodic tones can validate the assumption between perceived and actual ratings.
(click image to enlarge)
Figures 2 and 3 show the scores for the low-, medium-, and high-criticality tones for each of the two classes, as well as separate ratings for the perceived criticality (to validate the assumption of the correlation between the perceived criticality and the ratings). These figures show the following interesting points:
• The pulse tones show more criticality than the melodic tones, and the pulse tones also show a larger difference between the low-, medium-, and high-criticality alarms.
• The largest difference between the pulse and melodic alarms is in the activity dimension, which might indicate that this is the area that most strongly influences the ratings of criticality.
• The ratings for evaluation and power were as expected (slightly bad and increasing in power). The power dimension had minimal influence on the perceived criticality, however (see Figure 2).
The study shows the value and flexibility of this technique. In addition to providing overall ratings, the different dimensions and scales can provide targeted information about how to change the design to improve the overall scores.
Figure 3. Pulse tones show a significant difference between low, medium, and high criticality when the ratings are plotted for each attribute.
Product developers can implement these techniques quickly with a considerable degree of efficacy. The semantic scales that provide information about the three key dimensions people use to evaluate products can provide important direction for design. The scales help match consumer reaction to the designers' technology knowledge and creativity.
Getting the full benefit of the technique, however, can be realized only with an understanding of scale selection. Involving the entire design team in the interpretation of the data also leads to better application of the data. There are several key points to remember.
Customer Reactions Matter. It's important to let customers express what they want, but relying on the user's ability to voice features and appearance needs has limits. Relying on their ability to articulate features fails to address the experience that will resonate with consumers. Although users often do not know what they want from a design perspective, it's fundamentally important to gain insight as to how they want to react to a product. Using these scales does not make the user the designer, but it gives them a tool to communicate what they prefer.
Understanding Multiple Dimensions. While any random collection of scales provides insight into user preferences, it is important to use scales that reflect the multiple dimensions people use to evaluate a product.
Design Direction Is Enhanced, Not Substituted. This technique does not take the design process out of the hands of designers. Users are not asked to describe colors, shapes, form, or other characteristics with the scales. Rather, they are noting their desired reaction to a product. It is still up to the design team to create a design that leads to these reactions.
Reducing Subjective Bias. This technique also does not ask users which design they prefer. The response to preference questions can be influenced by many things: brand perception, previous experience, testing bias, value ratings, etc. This technique allows users to indicate their perceptions and expectations without the subjective bias of overall preference.
Repetition. It's important to repeat these exercises throughout product development to determine how well the design iterations are reaching the desired ratings (closer to the ideal or to a competitor, for example). The ideal should be reevaluated often. As other products come onto the market and technology innovations occur, users' ideal product ratings can change. It is critical to ensure that you're not designing toward an outdated ideal.
1. CE Osgood, GJ Suci, and PH Tannenbaum, The Measurement of Meaning (Urbana: University of Illinois Press, 1957).
2. WT Tucker, Experiments in Aesthetics Communications (Unpublished doctoral dissertation, University of Illinois, 1955), 68–70, 291–295.
3. D Coates, Watches Tell More than Time: Product Design, Information, and the Quest for Elegance (New York: McGraw Hill, 2002).
4. CE Osgood, WH May, and WS Miron, Cross Cultural Universals of Affective Meaning (Urbana: University of Illinois Press, 1975).