Enhancing Self-Design: Using Social Feedback to Lower Abandonment Rates

Imagine designing a pair of sneakers and having them delivered in a few days. Studies have shown that compared to “off-the-shelf” products of similar quality, customization leads to better preference fit and greater identification with a product. Self-designing products with co-creation configurators allows consumers to obtain the exact product they want—no compromise. Self-designing is fun, and consumers experience pride and feelings of accomplishment when they are part of the process. As a result, they have a high additional willingness to pay.

But there’s one big problem with this: Abandonment rates are perplexingly high. Data from 32 million customers across 18 countries shows that consumers prematurely abandon 95% of co-creation processes. Similarly, in an analysis of 15 German business-to-consumer configurators in categories as diverse as jewelry, spices, candles, and lingerie, abandonment rates ranged from 88 to 99%.

In a new Journal of Marketing study, we explore this contradiction between theory and practice. We measure individuals’ momentary emotional state continuously and unobtrusively on the basis of their facial expressions. Across six experiments where participants used different real configurators in different product categories, we employed FaceReader technology and an array of methods such as think-aloud protocols, automatic sentiment analysis, and survival analysis.

To quantify the problem, we conducted a survey of 771 American consumers who had abandoned self-designing a product with an online configurator at least once. We instructed them to think of the last time they abandoned such a process in each category and asked them to reveal their initial intentions (such as “before starting the configuration process…I had considered buying a self-designed [type of product]”). We collected information on 1,494 abandoned self-design processes. In 650 cases, consumers had been “shoppers” with no real interest in buying the product from the outset (44%); thus, their abandonment did not mean lost sales. However, in 844 cases (56%), consumers had at least considered buying the product when starting. Which bring us to the important question: Why do so many consumers leave prematurely when the academic evidence presented thus far suggests customization is supposedly so rewarding?

Beware the Valley of Death

Our findings show that consumers’ experiences are an emotional rollercoaster when self-designing. At first, expectations are high, which is why consumers begin the self-design process in the first place. However, their optimism and positive affect decreases quickly and they realize, to their frustration, that their interim design solutions are far from perfect and that the process of self-designing is less enjoyable than anticipated. Over-optimism gives way to under-optimism and consumers are not aware that they will learn more about their preferences and be more familiar with the configurator if they persist and continue self-designing. This momentary frustration and shortsightedness lead many to abandon the customization process prematurely. The perception that they have already invested some effort into self-designing does not create enough of a “lock-in” to prevent them from stopping the process. It is the consumers who overcome this “valley of death” in the U-shaped curve who gain the potential value from self-designing and regain a positive emotional state.

When participants successfully completed the self-designing process, they looked back on both the outcome and process positively. Which tells us that self-designing does indeed create high value for consumers—but only after completion.

Lessons for Chief Marketing Officers

What can companies learn from this study? How can they change the behavior of self-design configurators to reduce the abandonment problem?

Consumers who get positive social feedback (e.g., the short message, “Wow. Great design. Looks really good. I like it a lot,” by another consumer who is online) in the critical phase after a few minutes show significantly lower abandonment rates. However, positive feedback that is automatically generated has no such effect, no matter the form in which it is delivered.

Companies can think of integrating an online community to their self-design configurator. The added risk is that not all real peer feedback will be positive, potentially flipping the positive effect. Therefore, companies might consider hiring consumers for their benevolent feedback.

Companies might consider the possibility of simulating real peer consumer feedback. Rapidly progressing generative artificial intelligence possibilities make this a tempting option, although a number of ethical issues will warrant consideration before deciding to take this course.

Future research can potentially investigate additional ways of helping consumers overcome the critical phase in self-design processes. One possibility would be to provide attractive starting designs that only need to be refined and adapted rather than starting from scratch. If these starting solutions somewhat match consumers’ preferences, consumers may start with a design they already like and be more likely to complete it.