In a previous post we were lucky enough to get Frank Piller to tell us how he got into mass customization and what developments he as seen during his time researching the subject. In this post Frank reveals insights into where mass customization it’s advantages and drawbacks.
What are the most common shared factors in mass customization business models that fail? Is there one thing you would recommend a start up to avoid?
I believe many mass customization start-ups fail for the same reasons other start-ups fail: Lack of financing, inexperienced management, or just bad luck. What I can recommend for mass customization start-ups is, first of all, really decide where your system provides customer value. This may sound simple, but I saw too many start-ups that were build on the promise “when we customize, they will come”. Think of BookTailor, a site where users could customize travel guide books. Sounds like a great idea, but when you have to become your own editor of a guide book, you loose the most important value of a good travel guide: To think of the unexpected, to surprise, and to provide input in a situation not planned (say, an entire week of rain). BookTailor considered customization as a value per se. But this is not true. Customization is just the vehicle for customer value, but not its origin.
A second advice is to see whether your mass customization business is scalable once success will come. Often, mass customization is based on large work shops or sample room operations. But when you want to succeed with mass customization, you have to have stable processes for each customer order.
How important do you think social networking and the sharing of ideas/designs are? And is the building of an online community and the implicit exchange of social capital factored into many mass customization business models?
This is a rather new development, and I do not know too many communities or social networks that support a mass customization business (often, Threadless is named as an example, but Threadless explicitly is NOT mass customization, but a mass production model based on customer co-creation in the design stage.)
But we just started a 5 Million Euro research project called SERVIVE, funded by the European Commission, where our mission is to scale-up mass customization in the European fashion industry. One of the measures to do so could be communities supporting the configuration process. But you have to ask me again in three years if we succeeded.
The use of online ‘configurator’ software seems to currently be the most common format for mass customization, what do you see as the benefits and drawbacks of this model and how do you see this evolving in the future?
Yes, that’s correct. It was the broader development of online configurators that made mass customization happens in a larger scale. Have a look at our web-site for the scale and scope of configurators today.
The core drawback of most configurators, however, is that they are still parameter (option) based. Customers have to make their own decisions out of a list of pre-defined options. This often demands a large number of decisions and also knowledge of the user about the product. While this may be perfect in the business-to-business context where configurators originated, in consumer markets this is not always the best option.
Here, need-based configuration often is better. This means that users have to tell something about her preferences, requirements, or expected outcomes. This input then is transferred by an algorithm into a product configuration. There is a great paper by three scholars that compared the use of a parameter versus need-based configurator for Dell (asking people what graphic card they want versus asking people what games they play). In this paper, the authors clearly find that most users prefer the need-based solution, mimicking the behavior of a good sales person (T. Randall, C. Terwiesch, and K. Ulrich, User design of customized products. Marketing Science, Marketing Science, 26 (2007) 2 (March-April): 268-280) (here is a link to a previous publication). Here, I believe, industry has to invest much more in developing better configuration systems that minimize “mass confusion”.
Are there alternatives to mass customization?
Absolutely! I recently see better matching-systems for standard products as a strong alternative to mass customization. Within an assortment (of pre-fabricated products), customer specific choices/options are recommended. Consider My Virtual Model, a matching service for fashion retailers and the appliance industry. MVM enables consumers, either on its own site or on the sites of its clients, to build themselves in a virtual model (an avatar), by selecting different body types, hair styles, face characteristics, etc. Consumers also type in their basic measurements so that the virtual model represents their body measurement. In addition, customers can specify what kind of “fit” they prefer (loose, comfort, tight, etc.) so that the recommendations provided do not only fit the customer in terms of sizes and appearance, but also in terms of how they do feel inside the garment.
When MVM started offering virtual avatars in 1999, they looked more like a curious oddity. But now their avatars are used by more than 12 millions individual users. Companies such as Adidas, Best Buy, Levis, Sears and H&M are using these virtual models to generate business and stronger ties to their customers, lured by the increase in such metrics as average order value and conversion.
Any other example of such a matching service?
Sure. A great example is Zafu.com. Finding the right size of a pair of jeans is a challenge for many women. The answer of mass customization is taking a customer’s measurements and making a custom pair of jeans for her. Zafu offers a different approach. From the customer perspective, the experience starts similarly. Zafu asks women shoppers eleven questions about how they prefer jeans to sit on their hips or waist to create a body profile. In addition, they ask for some basic body measurements.
But instead of using this information to create a custom cut, they match it with a large database of proprietary fitting information about the jeans of more than 30 major brands. This database contains hundreds of styles, from broadly marketed Gap to pricey designer labels. The consumer then gets a list of ranked results, linked with the brand’s website to purchase.
Zafu’s personalization service is an alternative model to conventional mass customization. It may not have the inventory advantages and value prepositions of mass customization, but is much easier to implement and is a much faster scalable system. For consumers, such a matching service also implies less waiting time as well as no price premiums associated with custom products.
But both models supplement each other: For most consumers, a better matching service like MVM or Zafu will provide sufficient value. For others, however, the ultimate product still will be the truly custom jean — providing not only perfect fit, but also the hedonistic satisfaction connected with a custom product. Zafu is well positioned to profit from this trend. The company is owned by Archtetype, a major enabler of true mass customization for the clothing industry. Thus, they easily can refer a customer finding no fitting piece in Zafu’s database of the existing assortment of standard products to the custom clothing offerings.
I predict that we will see many more examples of these matching services as they offer companies to profit better from what they already have: vast assortments of existing goods. The result may be a new understanding of mass customization, beyond its roots in on-demand manufacturing and product design. In the end, it is the customer who drives the business. And customers are not differentiating between personalized, customized, or standardized offerings. I believe that we will need a broader understanding of mass customization. And I am excited to work on this challenge in the coming years.
Thanks again to Frank for sharing his views on mass customization, for more check out his configurator database and his Mass Customization, Customer Co-Creation & Open Innovation.