786Influence of Product Complexityand Customer Demographics on Co-DesignbyMobeen ShaukatA dissertation submitted in partial fulfillmentof the requirements for the degree ofDoctor of Philosophy(Design Science)in The University of Michigan2012Doctoral Committee:Professor Richard D. Gonzalez, Co-ChairProfessor Kazuhiro Saitou, Co-ChairProfessor Panos Y. PapalambrosProfessor Venkatram Ramaswamy

To my fatherShaukat AliWho was my first mentor, best friend, and everythingMay Allah Bless his soul million times a momentii

AcknowledgementsAll praise and praising is for God who makes miracles to happen. Thisdissertation is indeed one of them. Six years ago when I started as a doctoralstudent in Manufacturing Engineering, I had no idea that I will graduate in DesignScience instead and that these six years will be the roller coaster ride of my life. Ifaced number of challenges during these years and at times came close to givingup. But few dedicated individuals helped me to keep going and finish mycherished goal of obtaining a doctorate. I would like to take opportunity andexpress my gratitude for their guidance and assistance. First of all I wish to thankPanos Papalambros who helped me to get transferred to the Design Scienceprogram, arranged for the financial assistance when I needed it the most,assisted me in forming the dissertation committee, and solved so many otherissues that came during the course of the program. He went way out of his wayto help me to go through this program. Thank you so much Panos. Specialthanks are due to my advisors Richard Gonzalez and Kazuhiro Saitou for theirguidance, patience, encouragement, and valuable feedback. I would also like tothank John Marshall, Anocha Aribarg, and Tim Simpson for insightfulconversations that led to research questions presented in this dissertation. I amalso indebted to Venkatram Ramaswamy, a globally recognized consultant onco-creation, who despite his extremely busy schedule agreed to serve on thecommittee and provided unique perspectives about co-creation and co-design.Thanks are also due to JoAnn Heck and Scott Slagle who helped me to obtainresearch data from an online panel of customers. I would also like to thank Atiffor help with editing and proofreading. Finally I would like to thank my mother andsister for their encouragement and prayers for all these years.iii

Table of ContentsDedicationiiAcknowledgementsiiiList of FiguresviiList of TablesviiiAbstractixChapter 1 – Co-Design – An Introduction11.1 Co-creation and Co-design11.2 Co-design, Product Design, and Innovation51.2.1 Co-designed Products51.2.2 Co-design and Product Failure151.2.3 Co-design and Innovation161.3 Dissertation Layout18Chapter 2 – Literature Review and Research Objectives2.1 Literature review20202.1.1 Three Methods for Customer Involvement222.1.2 Co-design vs. Mass Customization and Personalization252.1.3 Experiments for Customer Involvement in Design282.2 Gaps and Challenges in co-Design Research312.2.1 Research Objectives and Questions33Chapter 3 – Influence of Product Complexity and Frequency of Useon Co-design3.1 Method343.1.1 Selection and Classification of Products353.1.2 Participants37iv34

3.1.3 Procedure383.1.4 Hypotheses393.2 Results403.2.1 Product Complexity and Frequency of Use413.2.2 Extraction of Design Ideas423.2.3 Ratings of Design Ideas453.2.4 Statistical Analysis473.2.5 Evaluation of Hypotheses503.2.6 Analysis of Novelty513.2.7 Role of Gender and Generation in Co-design543.3 Conclusion and Limitations55Chapter 4 – Customer Demographics and Co-design574.1 Procedure584.2 Results and Discussion584.2.1 Co-design and Product Type594.2.2 Co-design and Product Features644.2.3 Awareness of Co-design and Related Terms684.2.4 Idea Submission by Customers714.2.5 Idea Submission Medium734.2.6 Interaction Medium754.3 Conclusions and Implications78Chapter 5 – Implementing Co-design through Engagement Platforms5.1 Introduction80805.1.1 Design Thinking815.2 Fives Steps for Implementing Co-Design835.2.1 Invite855.2.2 Interact875.2.3 Ideate895.2.4 Implement915.2.5 Improve935.2Comparison between Chapter 1 framework and the 5i modelv95

Chapter 6 – Conclusions and Future Work996.1 Summary996.2 Contributions1006.3 Future Work102Appendix – Co-design surveys104111Referencesvi

List of FiguresFigure 1.1: Co-design at Lego8Figure 1.2: Co-design at Threadless9Figure 1.3: Co-design at Ponoko10Figure 1.4: Co-design at Muji11Figure 1.5: Co-design at Elephant Design12Figure 1.6: Co-design at Local-Motors14Figure 2.1: Co-design and various disciplines21Figure 2.2: Relationship between co-design, personalization27Figure 3.1: Verbal input as received from customer43Figure 3.2: Mean and standard for quantity of ideas48Figure 3.3: Mean and standard for novelty of ideas49Figure 3.4: Mean and standard for feasibility of ideas49Figure 4.1: Customer interest in co-design of various products60Figure 4.2: Customer interested in co-design by product type and gender62Figure 4.3: Product type and generation63Figure 4.4: Customer interest in co-design by product features65Figure 4.5: Customer interested in co-design by product features & gender67Figure 4.6: Product features and generation67Figure 4.7: Awareness of co-design terms69Figure 4.8: Idea submission medium.73Figure 4.9: Customers preferred interaction methodology.76Figure 5.1: Three spaces for a design project and role of designer.82Figure 5.2: Five steps for implementing co-design through engagement84Figure 5.3: Dell’s Ideas Storm home page for submitting ideas.87Figure 5.4: My Starbuck Idea home page.91Figure 5.5: Mapping between Chapter 1 framework and 5i model96vii

List of TablesTable 1.1: List of example companies using co-design7Table 3.1 Four types of products36Table 3.2 Statistics of subjects participated in survey37Table 3.3: Customers’ rating of product complexity41Table 3.4: Customers’ rating of frequency of use42Table 3.5 Extraction of design ideas43Table 3.6 Rating of design ideas46Table 3.7 Correlations and Cronbach Alphas for idea ratings46Table 3.8 Averages and S.D. for product quantity, novelty47Table 3.9 Comparison of products for quantity, novelty48Table 3.10 Percentage of novel ideas and most novel ideas52Table 3.11: Results of Ordinary Linear Regression54Table 4.1: Chi square values for customers’ interest in co-design63Table 4.2: Chi square significance values for product features68Table 4.3: Awareness of co-design terms by gender70Table 4.4: Awareness of co-design terms by generation70Table 4.5: Idea Submission by Gender and Generation72Table 4.6: Idea submission medium by gender.74Table 4.7: Ideas submission by generation.75Table 4.8: Interaction medium by gender.77Table 4.9: Interaction medium by generation.78Table 5.1: 5i model applied to co-design activities of companies97viii

AbstractCo-design is an emerging trend in new product development that results in moreactive customer participation in product development than conventional designprocesses; it connects customers, designers, and engineers throughout productdesign and development. This dissertation studies the influence of productcomplexity, frequency of product use, and customer demographics on co-design.Three co-design surveys were conducted with an online community ofcustomers. More than five hundred community members participated in thesesurveys and provided about two thousand design ideas to improve four differentproducts. The analysis of these ideas showed that frequency of use influencesthe quantity of design ideas generated for that product. It was also found thatproduct complexity (estimated by number of components and user interfaces)does not affect the novelty of design ideas generated for a product. In additionthe data suggests that customers’ dissatisfaction with existing products may leadthem to generate novel ideas during co-design.Results from these surveys also showed that customers’ interest in co-designvaries significantly with product type. For example customers are five times moreinterested in co-designing a house than an inkjet printer. Gender also influencesco-design. Females are more interested in co-designing clothes than maleswhereas males are more interested in co-designing cars than females. Age alsoplays a role in determining customers’ interest in co-design.Finally, a five step framework (5i model) was proposed to implement co-design.These five steps were identified after reviewing latest case studies of co-designefforts for various companies like Dell, Starbucks, and Apple. In addition, theix

data collected as part of this dissertation also informed the five step framework.The five steps are: invite the customers, interact with them, ideate for newproducts and services, implement customers’ ideas and then improve theprocess of co-design. Using these five steps a company can initiate co-designefforts and engage customers in product design and development. It is predictedthat in coming years co-design will increasingly augment conventional designprocesses.x

78678686CHAPTER 1CO-DESIGN – AN INTRODUCTION1.1 Co-creation and Co-designIn an increasingly competitive global economy, companies are motivated tocontinuously search for innovative products and services. Innovation has beenlong regarded as the source of profitable growth for companies. Customers canplay a pivotal role in innovation through co-creation. Co-creation is defined as thejoint creation of value by the company and the customer (Prahalad &Ramaswamy, 2004). Co-creation is a new approach for interacting withcustomers in various value creation activities undertaken by a company. In theearly 2000s, Coimbatore Prahalad and Venkatram Ramaswamy from theUniversity of Michigan’s Ross School of Business identified the co-creation trendin industry and brought it to the attention of academia. As they write in their book,The Future of Competition: Co-Creating Unique Value with Customers:1

“The meaning of value and the process of value creation are rapidlyshifting from a product- and firm-centric view to personalized consumerexperiences. Informed, networked, empowered and active consumers areincreasingly co-creating value with the firm.” (Prahalad & Ramaswamy,2004).According to McKinsey Quarterly, co-creation is one of the eight importantemerging trends for businesses across the globe (Manyika et al., 2008). Cocreation may also have indirect benefits such as increasing customerengagement, loyalty and interaction with the company, which may have positivelong term consequences beyond the immediate co-created outcomes.Co-design is a special instance of co-creation (Sanders, 2008). Co-designhappens when customers actively participate in the design and developmentprocess of new products. For this dissertation following definition of co-design isadopted:“We use co-design in a broader sense to refer to the creativity ofdesigners and people not trained in design working together in the designdevelopment process.”(Sanders, 2008).Co-design is different than conventional approaches to design. In conventionaldesign customer participation is usually limited to product specificationdevelopment through market surveys. They rarely participate in the initial phasesof the design activity where ideas and concepts are generated by companydesigners. The product is given to the consumer with the company perhaps2

offering variants of the product to coincide with market segments. However, inco-design the user or the customer becomes an active member of the designteam and participates in the design process. Therefore co-design results in usergenerated designs or the designs that are realized with active participation fromusers. During co-design customers can participate in concept design, conceptselection, and/or detailed design. The mode and intensity of participation varydepending on the product type and the design process.Several researchers have analyzed the role of customer involvement in productdesign. Eric von Hippel, a professor of innovation at MIT Sloan School ofManagement, studied the role of lead-users in product innovation. Lead usersare those customers who are at the leading edge of the marketplace. They tendto experience needs ahead of the other customers and they are able to innovatefor themselves (von Hippel, 1986). Working with these users can become asource of innovation for new products for a company. Von Hippel states in hisbook, Democratizing Innovation:“Users of products and services both firms and individual consumers areincreasingly able to innovate for themselves . Users that innovatecan develop exactly what they want, rather than rely on manufacturers toact as their (very often imperfect) agents.” (von Hippel, 2005)This quote illustrates that firms at times cannot anticipate customers’ needsaccurately. As a result the solutions (products or services) provided by the firmmay not be what customers want. In such cases, involving customers in thedesign process can lead to innovative products that customers want, desire, andperhaps more importantly, are willing to buy. In doing so, customers become3

active participants in the design process - usually considered to be the realm ofcompany designers and engineers.Along the same lines, Kaulio (1998) has identified three product developmentapproaches: design for the customer, design with the customer, and design bythe customer. The first approach relies on traditional market research methods,the second approach maintains a dialogue with the customer during the designprocess by letting them respond to various designs, whereas the third approachallows user to become the designer and this final approach is where co-design isrealized. Pals et al. (2008) has used the terms no direct user involvement,reactive user involvement, and active user involvement to represent the abovementioned three approaches to product design.Most current products are designed using the first two approaches. Usuallycustomers play a passive role in design through surveys and focus groups. Thatis, they typically respond to the designs already developed by companies.However through co-design customers are able to propose designs and becomeactive participants in the product design process. Such a co-design process hasseveral advantages over the conventional process. The customer provides thedesign and the customer makes a commitment for purchasing the product. In thisway no forecasting is required for the product volume. So every product made isalready sold out before production. Several companies have used co-design todesign successful products. Next section describes some current examples inindustry illustrating the use of co-design and its principles. A brief discussion4

about the influence of co-design in reducing product failures and innovation isalso given.1.2 Co-design, Product Design, and Innovation1.2.1 Co-designed ProductsA number of companies—big and small—are using co-design in productdevelopment process. These companies are listed in Table 1.1. Some of thesecompanies are supplementing conventional design activities with co-design whileothers co-design their entire product line. Currently, products designed using codesign mostly consists of toys, furniture, apparel, appliances, etc. A briefdescription of the co-design activities undertaken by these companies is given inthis section.In order to organize and compare the co-design activities undertaken by variouscompanies, a common framework is needed. To develop such a framework, theprocess from conception to consumption of a co-designed product is divided intosix steps: get design tool, design product concepts, submit design concepts,select design concept, execute detailed design & manufacture, and buy/sellproduct (see Figure 1.1 for an example with Lego). Most of the companiesrequire that a customer submits the product design in a specific format. Thisformat can be a CAD file or a design template. Some companies also accepthand sketches and photographs of the product to be co-designed. After obtainingthe design tool (a simple CAD package or a design template), the customer5

designs the product. This is the major distinction between co-design andconventional design as the customer—instead of a company designer—isdesigning the product. Based upon the product type, customers can either submitconceptual design and left the detailed design activity for company designers orin some cases customers submit detailed design that requires minimal input fromcompany designers.After completing the design, customer submits it to the company such as throughthe company website. Hence, the Internet is one of the enabling technologies forco-design. Next concept selection takes place. Two methods have been typicallyused: Either other customers on the company website vote for the best design orthe customer himself/herself decides to manufacture the design based upon thefeedback provided by the company. During concept selection stage customersalso show their commitment to buy the product. Most of the products that arecurrently being co-designed are simple and do not require detailed design (e.g.,toys, t-shirts). However, in some cases if detailed design is required (e.g.,vehicles), then company designers and engineers can undertake this step. Theproduct is then manufactured by the company or a contract manufacturer. Oncethe product is manufactured, it is shipped to those customers who showedinterest in buying the product. At times co-designed products are available forother customers (who did not participated in design) to buy as well.6

Table 1.1: List of example companies using ant design6Local MotorsProductsToysT-shirtsFurniture, household items, jewelry, miscellaneousFurniture, household itemsFurniture, home appliances, fashion items,miscellaneousVehiclesLego ( is a leading toy manufacturer with a customer base of more than 400million. It mass-produces toys that are designed by the customers (Hara, 2007).To help customers co-create Lego toys, the company has created softwarecalled Lego Digital Designer. Customers design Lego products using thissoftware and submit their designs to the company. Lego then selects theoutstanding designs and markets them. Lego shares the profits with thecustomer and also prints the customer name and picture on the Lego box (Hara,2007). In this example of co-design, the customer designs the products but thecompany makes the final decision about manufacturing a particular design. Thiscan be contrasted with the next example of co-design where voting is used for7

concept lDesignerDESIGNCustomerdesignsLego igns onLegowebsiteLegoselects thedesign.DETAILEDDESIGNANDMANUFA‐CTURENo detaildesignrequired.Legomanufact‐uresblocks viainjectionmoldingand shipsthem tothecustomer.Customerassemblesthe blocks.BUYOthercusto‐mers canbuy theproductfrom Legowebsite.Figure 1.1: Co-design at LegoThreadless ( has used designs submitted by its customers to build up a successfulbusiness. On its website, Threadless asks customers to submit T-shirt designs.The community of Threadless customers then votes for the best designs. Thedesigns that are voted best are printed on t-shirts that are available for purchaseonline or through a bricks and mortar Threadless store in Chicago. Threadlesspays fees to the customers whose designs are printed on the T-shirts. (Manyikaet al., 2008). It is conjectured that some of these customers may be designers byprofession in other fields and they use Threadless as an avenue to express theircreativity and design clothing that they want to wear. In this way Threadless is8

co-designing T-shirts where customers design the shirts and Threadless marketsthem. Here voting is used for concept selection.GETDESIGNTOOLCustomerdownloads templateto submitt‐shirtdesignsDESIGNCustomerdesigns gns onThreadlesswebsite.Threadlesscommuni‐ty votesfor adlessprints thevoteddesigns onT‐shirsts.BUYCustomercan buy t‐shirtsthroughonline andretailthreadlessstore.Figure 1.2: Co-design at ThreadlessCameseteria- a Brazilian T-shirt company is co-designing T-shirts in a similarway as Threadless. To date Brazilian customers have co-designed more than25,000 T-shirts on Cameseteria website. Customers can post their designs onthe website and they can also vote for the best designs, which are then producedby Cameseteria. Through co-design, Threadless and Cameseteria has reducedthe cost and risk involved in launching new apparel. On the cost side, co-designeliminates the expense of hiring professional designers and through customervoting it reduces the risk of launching those products that customers are notinterested in buying (Ramaswamy and Goulliart, 2010).9

Ponoko ( is an online co-design company. On Ponoko’s website customers cansubmit designs for various products like lamps, jewelry, furniture etc. Ponokomanufactures the products (or use contract manufacturers) according to thedesigns submitted by customers. The customer can sell the product at Ponokowebsite to other customers. One limitation on designs is that they have to bemanufactured using laser cutting process. Even with this limitation hundreds ofunique products are designed by the customers and are available for salethrough Ponoko’s website. In the Ponoko example, customers make the finaldecision about manufacturing the product. It is different than the previous twoexamples given where in one case the company made the final call and in theother case voting was used for concept eto ELECTCONCEPTCustomersubmitdesigns onPonokowebsitesPonokoprovidesthe pricingandprototypeto thecustomer.Customerdecides toorder theproduct.DETAILEDDESIGNANDMANUFA‐CTUREDetail eddesign Customercan fac‐tures theproductand shipsit to thecustomer.Figure 1.3: Co-design at Ponoko10

Muji ( of the most successful products sold by Muji, a Japanese retailer, aredesigned by Muji customers. Muji customers submit product ideas on its website.Muji customers also shortlist the ideas submitted on the website through voting.Company designers conduct the detailed design for the product ideas shortlistedby the customers. Those products are then manufactured and marketed by Muji.Products co-created by Muji’s customers tend to outperform other Muji productsdesigned using conventional design processes (Ogawa & Piller, 2006). Forexample a co-designed bean bag sofa generated sales of 1,344 million Yen ascompared to 24 million Yen of average annual sales in this product category.(Ogawa & Piller, 2005). A combination of voting and company designers’ opinionis used for concept selection for hesand/ordescribesfurnitureor bmitdesigns onMujiwebsiteMujicustomersvote forbestdesignsand showcommit‐ment omersbuy nufac‐tures theproduct.Figure 1.4: Co-design at Muji11

Elephant Design ( Design -- a Japanese design consulting firm -- claims that the user canorder anything they want by submitting a design idea on their website (ElephantDesign, 2008). The design process starts with a user submitting a product ideaon the website. Website users can view all the submitted ideas and they vote forthe best ideas. If enough users vote for a product design and they are ready topurchase the design at a given price, then Elephant-Design contacts amanufacturer to make that product. The detailed design phase is eitherconducted by Elephant-Design designers or the contract manufacturer’sdesigners. A compact electric cooker has been designed using this process. LikeThreadless, voting by customers is the primary tool for concept selection atElephant tchesand/ordescribesfurnitureor bmitdesigns onElephantdesignwebsiteElephantdesignusers votefor bestdesignsand showcommit‐ment omersbuy theproduct.Detaileddesign ctsthemanufac‐turers tomake theproductsvoted byusers.Figure 1.5: Co-design at Elephant Design12

Local Motors ( is applying co-design to automotive design. On Local Motors’website customers (most of them are freelance automotive designers) cansubmit their sketches for exterior design of a car. The site’s community thenvotes to select a design for production. To date only one design has beenselected for production named Rally Fighter. Even though the exterior designwas co-designed, most of the components for this car were selected off the shelf.Local-Motors engineers and machinists conduct the detailed design and fit theexterior design to a common chassis. The design is then transferred to a networkof suppliers that provides sub-assemblies to local factories. The car is thenassembled in small build centers where customers can participate in theassembly process.Defense Advanced Research Projects Agency (DARPA) has also initiated aprogram named The purpose of the program is to develop acollaborative environment that will allow co-design of complex electro-mechanicalsystems.13

signs onLocal‐Motorswebsite.Customersvote forbestdesignsand showcommit‐ment omersbuy theproduct.Finalassemblyis done inlocalshops.CustomerscanparticipateFigure 1.6: Co-design at Local MotorsCo-design has also been found helpful in eco- design and in designing low costproducts for developing countries. For example, Bhalla (2011) has described thathow Phillips Design co-designed a simple cooking stove to reduce CO emissionsand improve fuel efficiency. Phillips Design involved local NGOs and usedinsights from actual users to develop this simple cooking stove that has wonseveral design awards. Bhalla (2011) also suggests using co-design for a similardesign project aimed to design a 300 house for the world’s poor. Herecommends involving people who will live in those houses in the designprocess. Similarly Thrift (2006) argues that customers can influence companiesthrough co-design imposing their values on the design process. For example,customers can ask for sustainable products through co-design.This section presented several examples of co-design as it currently exists in themarketplace. Most of these examples can be conceptualized as providing theinfrastructure for individuals to manufacture and sell their designs. For example,14

an individual with an idea for a t-shirt can submit the idea and make use of anelaborate infrastructure to bring the idea to market. There is no need for anelaborate business plan. Potential customers vote on the idea, manufacturing ishandled through the infrastructure and sales are sometimes secured prior tomanufacture. Admittedly, these existing examples of co-design are somewhatlimiting. One wonders what co-design would look like for well-establishedcompanies such as Ford Motor Co or Wilson Sporting Goods. Some newerexamples of co-design have a somewhat different structure than depicted in sayFigure 1.1, such as the smart-phone landscape with Apple’s iOS and Google’sAndroid. This dissertation hopes to address these more general concerns of codesign motivated by the successful examples provided in this section.1.2.2 Role of Co-design in reducing Product FailuresCooper and Kleinschmidt (1987) have reported that about 35% of the newproducts failed to meet financial goals. Similarly, Ogawa and Piller (2006) havementioned that more than 50% of new products fail to meet financialexpectations. This implies that companies may in part be often wrong inpredicting what customers really want, at what price, etc. Any tool that can helpcompanies perform better at predicting customers wants, desires, and purchasebehavior will help reduce product failure rates. As Ogawa and Piller (2006) write:“The manufacturer’s nirvana is to develop and produce exactly what itscustomers want and when they want it and to do this with no risk ofoverstocks or inventory.” (Ogawa and Piller, 2006).15

Co-design provides one answer to this problem. Using a design process similarto Elephant-Design, companies can solicit customers’ suggestions about futureproduct offerings. The Internet has facilitated the communication requiredbetween companies and their customers. Companies can also judge customers’willingness to buy at given price and predict the volume to produce. In this waythe design process has been initiated by the customer and customers may showa greater commitment to buy the products once designed.1.2.3 Co-design and InnovationAs stated earlier in the chapter that companies are on a continuous search forinnovative products and services. Through co-design, customers can bring theirunique experiences and play a pivotal role in innovation. Examples of productsdesigned on Elephant Design, Muji, and Ponoko websites demonstrate thatcustomers were able to design unique and novel products through co-design.However an important question is what type of innovation happens in co-design?A review of the literature shows two classificatio

Chapter 4 - Customer Demographics and Co-design 57 4.1 Procedure 58 4.2 Results and Discussion 58 4.2.1 Co-design and Product Type 59 4.2.2 Co-design and Product Features 64 4.2.3 Awareness of Co-design and Related Terms 68 4.2.4 Idea Submission by Customers 71 . Starbucks, and Apple. In addition, the .