emotional design; application of a research-based design approach

ORIGINAL PAPER

Emotional Design; Application of a Research-BasedDesign Approach

P. M. A. Desmet & R. Porcelijn & M. B. van Dijk

Received: 15 December 2006 /Accepted: 20 March 2007 / Published online: 11 September 2007# Springer Science + Business Media B.V. 2007

Abstract In this paper, we discuss an approach to‘design for wow’ that focuses on the emotions thatconstitute a wow-experience. In this approach, theeliciting conditions of these emotions are used todefine a product character with a high wow-impact. Inaddition to the approach, a measurable wow-index isintroduced. First, a concept of wow is described inwhich wow is explained as a combination offascination, pleasant surprise, and desire. The elicitingconditions of these three emotions are examined andcombined to a ‘wow-appraisal.’ This concept isapplied in the design of a mobile telephone. Acombination of qualitative and quantitative researchmethods was used to formulate concern themesrelevant for the wow-appraisal. These themes wereused to create a layered product character and a

prototype of the final design. An evaluation studydemonstrated that the product designed with thisapproach rated higher on the wow-index than alter-native products.

Keywords Emotions . Product character .

Wow-impact .Wow-index .Wow-appraisal

Introduction

Some products are more exciting than others, and thisexperience of excitement is often seen as something thatshould be strived for because it motivates customers toprefer one product over another (Wakefield and Baker1998). Consumer researchers have argued that thoseproducts that excite customers will be more successfulthan those that do not (Millard 2006). In line with thisview, Terninko (1995) proposed a model of differentialadvantage that distinguishes between three types ofproduct features: basic, performance, and excitementfeatures. Basic features are those that the product typeis expected to deliver. A basic feature of a mobilephone, for example, is a database function to storetelephone numbers. Performance features are those thatdifferentiate between competing products. Mobilephones have performance features such as weight,screen size, and stand-by time. In his model, Terninkoproposed a linear relationship between consumersatisfaction and performance features: the higher the

Know Techn Pol (2007) 20:141–155DOI 10.1007/s12130-007-9018-4

P. M. A. Desmet (*)Department of Industrial Design,Delft University of Technology,Landbergstraat 15,2628 CE Delft, The Netherlandse-mail: [email protected]

R. Porcelijn :M. B. van DijkKVD Reframing and Design,Westhavenweg 59f,1042 AL Amsterdam, The Netherlands

R. Porcelijne-mail: [email protected]

M. B. van Dijke-mail: [email protected]

performance, the higher the level of satisfaction. Thethird type of features, excitement features, are thosethat the consumer did not expect to see in the productand is excited to learn are there. GPS systems or high-resolution video cameras can be excitement features ofmobile phones. Following this feature-based excite-ment model, Mann (2002) proposed that those whowant to design products that evoke feelings of ‘wow’should find and apply one or more ‘excitementfeatures’.

The main difficulty of this approach is that thiskind of excitement does not last. A photo camera onmobile phones, which was an excitement featurewhen it was first introduced (“wow, I can takepictures with my phone”), quickly became a basicfeature of the next generation of mobile phones. Thefocus on excitement is basically a focus on surpriseby newness, and a feature that excites today willinevitably stop being exciting tomorrow – or the dayafter tomorrow. This technology-focused approachlacks direction apart from the search for newness. Inthis paper, we propose an alternative approach thatfocuses on the relationship between users and theirproducts. The approach is user-oriented rather thantechnology-oriented, and holistic in the sense that thewow-impact of the product is considered on aconceptual rather than on a feature level. The mainadvantage is that this approach generates a unifiedconcept in which (visual) appearance, interaction, andfeatures are aligned to elicit a coherent experience.First, the emotional structure of a wow-experience isanalyzed. On the basis of this analysis, we proposethat wow can best be considered a compoundedemotional response. Then, a design project is dis-cussed in which a mobile telephone was designed toevoke wow. The paper describes the design process,the results, the evaluation study, and discussesimplications and future research directions.

Product Emotion

Products can evoke many different kinds of emotions.We can admire the latest ultra-slim laptop, be irritatedby an annoying telephone, feel at ease in a comfort-able sofa, and so on. And although the touch ofmelancholy felt when coming across a long forgottenchildhood teddy bear seems incomparable to the thrillof driving a motorcycle, both these responses belong

to the wide spectrum of human emotions. Our studiessupport the idea that even conventional products, suchas mobile telephones and fabric softeners, can evokecomplex and layered emotional profiles (see Desmetet al. 2001). The last few years, we have beenexploring the possibilities to employ such emotionprofiles as the starting-point for new ‘emotionallyinnovative’ designs (see Desmet and Dijkhuis 2003).We found that by measuring emotional responses, itwas possible to design products that target specifictypes of emotions. For this project, it was decided tofollow a similar approach in which research was usedto feed the design process. Because an integratedapproach was believed to be constructive, both thedesigner (second author) and the researcher (firstauthor) were involved in all stages of the project.

Appraisal Theory

Most contemporary emotion theorists view emotionsas coherent, organized, and functional systems. Infact, the basic Darwinian presupposition that emo-tions fulfill some sort of adaptive function is probablyshared by all psychologists. In this view, emotions areconsidered the mechanisms that signal when eventsare favorable or harmful to one’s concerns. Theprocess of signaling the relational meaning of anevent is commonly conceptualized as ‘a process ofappraisal.’ An appraisal is a “direct, non-reflective,non-intellectual automatic judgment of the meaningof a situation” (Arnold 1960 p. 170) in which ourconcerns serve as points of reference. FollowingArnold (1960), Frijda (1986) argues that when weappraise a stimulus as beneficial to our concerns, wewill experience positive emotions and try to approachthis particular stimulus. Likewise, when we appraise astimulus as colliding with our concerns, we willexperience negative emotions and try to avoid it.Most contemporary researchers in the cognitivetradition of emotion hold that particular types ofemotions are associated with particular types ofappraisals, and that emotions can be predicted fromthe nature of the underlying appraisal and concern. Inour view, design for emotion requires an approachthat focuses on appraisals and concerns. Althoughemotions may be subjective, the process of emotion isuniversal and people that share concerns and apprais-als will experience similar emotions to a givenproduct.

142 P.M.A. Desmet, et al.

Product Concerns

Ortony et al. (1988) developed a typology of humanconcerns in which three types of emotional concernsare distinguished: goals, standards, and attitudes.This typology was used in the current projectbecause Desmet and Hekkert (2002) found thesethree types of concerns to be particularly valuable fordescribing concerns that are relevant for productemotion. The first concern type is goal. Goals areevent related concerns. These are the things onewants to get done and the things one wants to seehappen. The goals that people try to satisfy are oftenassumed to be structured in a hierarchy rangingbetween abstract goals or aspirations, like the goal tohave a successful life, and goals as concrete andimmediate as the goal to catch a train. Many goalsare directly and indirectly activated in the human–product relationship. For example, we buy, own, anduse products because we believe they can help us toachieve things (a digital agenda to make us moreorganized) or they fulfill a need (a bicycle fulfils theneed for transportation).

The second concern type is standard. Standardsare our beliefs, norms or conventions of how wethink things should behave. Whereas goals refer tothe state of affairs we want to obtain, standards arethe states of affairs we believe ought to be. Forexample, many of us believe that we should respectour parents, eat more fruit and vegetables. Moststandards are socially learned and represent thebeliefs in terms of which moral and other kinds ofjudgmental evaluations are made. Whereas goals arerelevant for our personal well-being, standards arerelevant for the preservation of our social structures(and thus indirectly also for our personal well-being). We approve of things that comply withstandards and disapprove of things that conflict withstandards. We not only have standards regardinghuman (inter)action, but also regarding products.With respect to products, our values represent ourbeliefs of how a product (or a person associated withthe product) should behave or function. We expect acar to drive, and we expect a vase to be water-resistant.

The third concern type is attitude. Our attitudesare concerns that are related to objects. Theyrepresent our dispositional likings or dislikings(taste) for particular (attributes of) objects. We have

affect dispositions towards product types (“I don’tlike microwave ovens”), towards aspects or featuresof products (“I like red cars”), towards style (“I likeItalian design”), towards quality of interaction (“Ilike cars that have a firm drive”), and towardscontext-related consequences of products (“I likefeeling relaxed after drinking a beer”). Some peoplelike red cars, others like black cars. Some peoplelike Italian furniture style, whereas others prefer theScandinavian style. Emotional responses related toattitudes are elicited by the appearance of the‘object as such’ and not by an (anticipated)consequence of usage or by (expected) behavior orfunctioning.

The Concept of Wow

What is a wow-experience? Essentially, “wow” isnothing more than a simple verbal exclamation.The research literature does not indicate attempts todefine or explain the conditions that elicit thisexperience that would enable us to be more precise.This is not surprising given that the concept ofwow belongs to the domain of everyday experi-ence, and is a laymen’s rather than a scientificconcept. To some extent, we all have experiencedwow, and thus, from experience we all know whatwow is. It is not difficult to come up with examplesin the domain of product design: “wow: that is thecoolest car”; “wow: look at the size of that coffeecup”; “wow: feel how this knife cuts the onion as ifit is made of butter”. On the basis of an explorationof these and similar everyday examples, weconsider wow to be an affective experience, thatis, a valenced experience that includes an unpleas-ant or (as in this case) a pleasant feeling. Moreover,we consider wow to be an emotional experiencebecause all four components that constitute emo-tions can be observed in wow-experiences: asubjective feeling, a facial and bodily expression,a change in arousal, and a motivational tendency(see Lazarus 1991). But if wow is an emotion, whyis it not mentioned in the emotion discourse found insocial sciences literature? Why is it not included inany of the reported emotion typologies or taxono-mies? Our explanation is that wow is not an emotionas such, but represents a composite of severalpleasant emotions: pleasant surprise, desire, andfascination.

Application of a research-based design approach 143

Wow Emotions

Many emotion models advanced to date includeappraisal themes to differentiate between emotions. Arelevant question is therefore what particular appraisaltheme evoke the wow-experience. On the basis ofreported appraisal themes, Desmet (2003a) developeda typology of emotion types that we generallyexperience in response to consumer products. Fiveof these types represent pleasant emotions, and threeof these five are considered to combine to a wow-experience: pleasant surprise, fascination, and desire.

Pleasant surprise We experience surprise in responseto products that we appraise as novel, both in terms ofsuddenness and unexpectedness. To experience pleas-ant surprise, the product must be appraised asunexpected and as suddenly matching a concern.

Fascination We experience fascination in response toproducts that we appraise as unfamiliar. If the producttouches on our need of curiosity, we will experience thepleasant sense of unfamiliarity that we call fascination.

Desire We desire for products that we appraise as fitobjects that call for possession. The product isappraised as a fit object because using or owning itwill fulfill a goal, its design is delightful or itexpresses some desirable quality.

If wow is a combination of pleasant surprise,fascination, and desire, the wow-experience should beelicited by products that are appraised as unexpected,unfamiliar, promising, and fit for possession. Thedesign value of this proposition was explored with acase in which a ‘wow’ mobile telephone wasdesigned, prototyped, and tested.

Case of Design for Wow

In 2003, one of the main Dutch telecom companiesinitiated a design project to develop a mobile telephonemodel. The new model was to serve as the visualidentity for their mobile Internet service. In addition,the company wanted the new design to evoke a wow-experience. The project consisted of three parts: anexploration, a design, and an evaluation stage. In theexplorative stage, two studies were preformed that

resulted in a concern profile. This concern profile wasused to develop a product concept in the design phase.In the evaluation stage, it was studies if the newproduct concept elicited the intended wow-response.

First Stage; Exploration

The aim of the design project was to design a mobilephone that the target group users appraise as unexpected,unfamiliar, promising, and fit for possession. For thatreason, two studies were performed to explore thisappraisal and the underlying concerns of the target groupusers. In the first study, emotional responses elicited byexisting mobile phones were measured. The results wereused to create an emotional benchmark. The benchmarkidentifies three current wow-telephones. The secondstudy was designed to understand the underlying userconcerns that are responsible for this wow-experience.The first study was a quantitative measurement, and thesecond a qualitative group discussion.

Emotional Responses to Existing Designs; Study 1

The emotional responses elicited by eight existingmobile phones were measured in study 1.

Stimuli

Eight mobile telephones were selected by design andmarketing managers of the telecom company whowere instructed to select telephones of which theybelieved had excitement features that evoke a wow-experience. Two additional criteria used to selectstimuli were (1) the telephones should be new to theDutch market and (2) they should show substantialdesign variation. Figure 1 gives an overview of theeight selected stimuli.

Models A, C, E, and G have ‘candy bar’ shapes(no sliding or folding mechanism). Model B slidesfrom a closed to an open position; models D and Fflip from a closed to an open position; model H twistsfrom a closed to an open position. All models arecolored silver and black, except for model D that iscolored metallic red, silver, and black.

Participants

Thirty-five participants (20 men, 15 women) between20 and 28 years of age (M=24.8; SD=2.3) participated.


They were students of Delft University of Technology,recruited from the university cafeteria, and paid fortheir participation. All participants were in possessionof a mobile telephone, and used their telephone fivetimes (or more) a day to call, and two times (or more) aday to sent an SMS. None of the participants weresubscribed to a mobile Internet provider.

Method

The product emotion measurement instrument(PrEmo, see Desmet 2003b) was used to measurethe emotional responses elicited by the stimuli. Thisinstrument was used because it was especiallydeveloped to measure emotional responses elicitedby product design. PrEmo is a nonverbal self-reportinstrument that uses animated cartoon characters to

represent emotions. Instead of relying on words,respondents can report their emotions with the useof these characters. The interface of PrEmo depictsstills of the characters, each accompanied with athree-point scale. Table 1 gives an overview of theemotions that were measured in study 1 with theaccompanying animation stills.

The scale represent the following ratings: “I do feelthe emotion”, “to some extent I feel the emotion”, and“I do not feel the emotion expressed by thischaracter”. During the study, respondents were showna mobile telephone and instructed to use the charac-ters to report their emotional response(s) elicited bythe model. The study was conducted in a neutralroom, participants participated individually, and theorder in which the models were presented wasrandomized over respondents.

Fig. 1 Stimuli study 1

Table 1 Emotions measured in study 1


Results

To obtain a graphical representation of the results, acorrespondence analysis was performed with twofactors: stimulus (eight levels; each level representinga mobile phone) and emotion (ten levels; each levelrepresenting a measured emotion). Figure 2 shows thetwo-dimensional solution of this analysis. The firstdimension, which is represented by the horizontalaxis, accounts for 65% of the total inertia, and thesecond dimension, which is represented by thevertical axis, accounts for 29% of the total inertia.

The solution in Fig. 2 shows a ‘mobile phone andemotion’ space that visualizes the associations betweenthe phones and reported emotional responses. The

distances between the stimuli reflect the relationshipsbetween these stimuli (with similar phones plottedclose to each other). Similarly, the distances betweenthe products and the emotions reflect the relationshipsbetween them. This means that products that areplotted close to each other elicited similar emotions,whereas those that are plotted at a distance form eachother elicited different emotions. The responses elicitedby models C and E, for example, are more similar thanthose elicited by models C and A.

The space indicates that three models elicit a higherlevel of wow-experience than the other five: models A,B, and H. These three models are positioned closest tothe three emotions pleasant surprise, desire, andfascination. Table 2 shows the mean ratings of all

Fig. 2 Mobile telephoneand emotion space, study 1


telephone models on these three wow-emotions (on athree-point scale rating ranging between one andthree). The last row, the ‘wow-index,’ shows theoverall mean rating on these three emotions.

The telephone models are ordered in accordancewith their wow-impact in Table 2. The model in thesecond column (model B) has the highest wow-impact, whereas the model in the ninth column(model G) has the lowest wow-impact. Consistentwith Fig. 2, the table indicates that models B, A, andH evoke the highest level of wow-experience. In ourview, this indicates that these three models match theconcerns that are relevant for the wow-appraisal. Asecond study was performed to understand theseconcerns that underlie the wow-appraisal.

Design Parameters; Study 2

The aim of study 2 was to identify the emotionalconcerns that are relevant in the process of experiencingdesire, fascination, and pleasant surprise in response tomobile telephones. For this purpose, two group dis-cussions with target group consumers were organized.

Stimuli

The same eight mobile telephones as in study 1 were alsoused in study 2. The telephone and emotion map thatresulted from study 1 served as an additional reference.

Participants

Ten participants (five men, five women) between 21and 28 years of age (M=24.3; SD=2.0) participated,

divided in two groups of five participants. The firstgroup consisted of two male and three femaleparticipants, and the second group consisted of threemale and two female participants. Participants wererecruited from the university cafeteria, and paid fortheir participation. All participants had the Dutchnationality and were in possession of a mobiletelephone. They used their telephone five times (ormore) a day to call, and two times (or more) a day tosend an SMS. None of the participants were sub-scribed to a mobile Internet provider, and noneparticipated in study 1.

Procedure

The relationship between product design and emotionalresponses to mobile telephones was discussed in 2-h sessions. The group members first discussed theemotional impact of each of the eight mobile telephones.The models were placed randomly on the table, and themoderator invited participants to express their affectiveresponses. After this first part that took place inapproximately 1 h, the moderator introduced andexplained the mobile telephone and emotion map. Heexplained the map and how it was produced. The groupmembers were stimulated to discuss to what degree they‘agreed with’ the map, that is, to what degree they feltthat the map represented their emotional responses.After that, the group was stimulated to discuss theunderlying reason why the models and emotions wereplaced in that particular configuration. By using atechnique that was loosely based on the depth interviewtechnique ‘laddering’ (see Reynolds 1986), the moder-ator was able to direct the discussion to the abstract

Table 2 Measured wow-impact in study 1


level of underlying concerns. All discussions wererecorded on audiotape.

Results

All comments made by any of the participants thatreferred to concerns were recorded and subsequentlycategorized in terms of the three concern types goals,standards, and attitudes. Table 3 gives an overview ofthe categorized concerns. The first column shows theconcern type, the second shows examples of men-tioned concerns, and the third column shows concernthemes that were formulated to represent the concernexamples.

The concern themes that are considered to berelevant for a wow-experience (given the target groupand the particular product) are:

Goals I want a mobile telephone that is manageable,practical, and reliable.

Standards A telephone should be of best quality andlogical in usage.

Attitudes I like telephones with designs that areconsistent, unique and luxurious.

New Product Design

The concern themes served as points of departure inthe design phase. The first step was to define aproduct character; the second step to translate thischaracter to a product concept; and the final step wasto make the detailed design.

Product Character

Some of the mentioned concerns in Table 3 appear tobe inconsistent or even conflicting. In the process ofdefining a product character, we found these appar-ently conflicting concerns to be particularly inspiring.For example, the telephone should be innovative,surprising, and stimulating, and at the same time, itshould be no-nonsense and harmonious. Or, theappearance should be simple and balanced, and atthe same time, distinctive and unique. These para-doxical concerns are interesting because they consti-

Table 3 Concern profile found in study 2

Concerntypes

Concern examples Concernthemes

GoalsI want atelephonethat

Fits my hand (not too small or too big); is convenient to store; does not have awkward foldingmechanisms.

Manageable

Has clear and unambiguous buttons; has buttons that are easy to operate; provides me with clearinteraction feedback; can be operated with uncomplicated interaction protocols.

Practical

Does not have parts that can break off; is not sensitive for damage; has a protected screen; hassturdy buttons; is always reliable.

Reliable

ValuesA telephoneshould

Have a solid cover; not be too light; have a clear click if it has a folding or twisting mechanism, nothave the tactile quality of plastic; not make cracking sounds when held in the hand.

Quality

Be recognizable as telephone; have a design that emphasises the telephone function; should befunctional; have a no-nonsense expression; not have ‘design-frills;’ should not have unnecessaryornaments.

Logical

AttitudesI liketelephonesthat

Are consistent in general shape, colour and buttons; are shaped geometrically; have even shapes; havesimple shapes; have balanced shapes; have quiet shapes; have harmonious shapes; have symmetricalshapes; are made of real materials like metal (instead of plastic with metallic paint) and rubber.

Consistent

Are not boring; have powerful shapes; show distinctive features; have distinguishing characteristics;have an innovative design; have an exciting design; are stylish; are elegant; are modern; are unique(not mainstream).

Unique

Have classy design; are made of beautiful materials; are well-detailed and finished; do not look childish;do not look cheap; show perfection in design and fabrication.

Luxurious


tute a tension field by defining a design space thatexcludes existing design solutions. On the basis of theconflicting concerns, a character triangle was createdthat represents the eight concern themes. Figure 3shows the character triangle in words and pictures.

This product character includes three successivelayers: the first impression is impetuous and self-willed;then the character becomes sincere and balanced; andfinally it becomes beneficent and sophisticated.

Product Concept

In the process of sketching product ideas, the designeraimed to create a product concept that fits the layeredcharacter. In addition, the initiating telephone compa-ny defined a set of predefined product design require-ments, such as features to be included (e.g., anInternet function, a photo camera), and restrictionsin the types of buttons that were to be used. The firststep was to decide on a structural solution, which is adefinition of how the three main functions (i.e.,

telephone, Internet, and photograph) are combined toa unified product concept. In study 2, it was foundthat the one of the concerns of the users is thestandard that the telephone function should bepredominant and direct in usage. It was thereforedecided not to integrate the telephone, photograph,and Internet functions in a single interface, but toseparate them. In the light of this decision, two designdirections were explored: structures in which eachfunction has an equal presence, and structures inwhich the telephone function dominates over theother functions. For both strategies, structural solu-tions were explored through sketching and modeling.Figure 4 shows some of the sketches that were madeto explore structural solutions.

Figure 4 shows six structures: equal dominancewith (A) single, (B) double hinges, and (C) doublesquared hinges; unequal dominance with (D) single,(E) double, and (F) double squared hinges. The equaldominance with double squared structure (E) wasselected because it was believed that this best fits the

Fig. 3 Product character that represents eight concern themes


character and concern themes. Equal dominance isrequired because that allows for a design that is‘honest’ in the sense that it directly communicates itsfunctions. A double hinge is required because thatallows for a clear separation between the Internet andphone function that will support a simple and directinterfacing. The hinges should be squared becausethis will generate a design that uses prototypicalconfigurations (a phone by two rectangles in decon-nected lengthways and an Internet ‘computer’ by tworectangles connected breadthways).

Product Design

The product design was explored with sketches andmodeling. To prevent a wow-response that has only ashort existence, it was decided to focus on the overall(holistic) concept rather than on feature-based con-cepts. Sketches explored the possibilities to create alayered response in which each character domainrepresents one layer of experience: an initial impact atfirst sight, a second impact, and a long-term impact.The first impression of the product is that it is

Fig. 4 Structural product solutions

Fig. 5 Design sketches


impetuous and self-willed; when using the productone experiences its beneficence and sophistication;and after a period of using the product one willexperience its sincerity and balance. Figure 5 showssome of these design sketches.

The final design has three functional layers. Eachfunctional layer is built from three material layers.The two outer layers are made of soft rubber and theinner layer is made of metal (see Fig. 6).

Figure 6 shows that the layers are curved and eachlayer fits in the other. The rubber layers includeinterfaces, buttons, and hinges. The metal layerscreate a solid structure. The three functional layersdefine three product functions as shown in Fig. 7.

Figure 7 shows the final design. When closed, theproduct looks like a photo camera, and when opened itlooks like a small computer, or like a simple mobilephone, depending on how it is opened. The parts thatare touched (by the finger when interacting with theproduct or by the face when using the telephonefunction) are made of white soft rubber. The exterior ismade of cool metal (aluminum anodized in the colorgold). The basic dimensions are: 122×42×22 mm.

The design has an impulsive spirit; the basic shapeis clear, but the lines are playful. The product showsdiversity in color and material (both in visual and intactile quality): soft rubber for all operating elements;plastic pieces for the construction, and metal piecesfor the protection. The shape (and therefore also thecharacter) changes when opened. The layered charac-ter is applied in all details of the product in order tocreate a subtle and sophisticated design experience.

Third Stage: Evaluation

An evaluation study was performed to assess if theproduct design elicits the intended wow-experience.A prototype was built of the new design. Thisprototype did not function as a telephone, but wasfunctional in terms of interacting with the physicalobject (opening and closing the telephone and

Fig. 6 Three functional layers

Fig. 7 New mobile phone design concept


computer function). Study 1 was repeated with thenew prototype as an additional stimulus.

Stimuli

It was decided that, similar to study 1, the evaluationstudy should include eight mobile telephones. Thefirst stimulus was the prototype of the new design asshown in Fig. 8.

The second stimulus was a prototype of anadditional design that was developed by anotherproject team. The remaining six models were selectedfrom the set of eight that was used in study 1. Thefollowing models were included: A, B, D, F, and H.Model G was not selected because in study 1 thismodel was found to elicit least wow of all models (seeTable 2). Model C was omitted because a repeated-measures ANOVA (with emotion as the dependents)

indicated no significant differences in emotionalresponses between models C and E. Figure 9 givesan overview of the eight selected stimuli.

All stimuli were nonworking prototypes. Models Aand E are ‘candy bar’ shapes (no sliding or foldingmechanism). Models B and C slide from a closed to anopen position; models D, F, and G flip from a closed toan open position; model H twists from a closed to anopen position. All models are colored silver and black,except for model D that is colored metallic red, silver,and black; model C that is blue and gray; and model Gthat is metallic gold.

Participants

Thirty participants (15 men, 15 women) between 20and 28 years of age (M=24.8; SD=2.3) participated.They were students of Delft University of Technolo-

Fig. 9 Stimuli study 3

Fig. 8 Prototype of the new design


gy, recruited from the university cafeteria, and paidfor their participation. All participants were inpossession of a mobile telephone, and used theirtelephone five times (or more) a day to call, and twotimes (or more) a day to sent an SMS. None of theparticipants were subscribed to a mobile Internetprovider, and none participated in study 1 or 2.

Method

PrEmo that was used in study 1 was also used instudy 2 to measure the emotional responses elicitedby the stimuli. The participant was shown a stimulusand instructed to use the characters to report his or heremotional response(s). The same procedure was usedto measure the emotional responses elicited by theother stimuli. The study was conducted in a neutralroom, participants participated individually, and stim-uli were presented in a randomized order.

Results

To obtain a graphical representation of the results, acorrespondence analysis was performed with twofactors: stimulus (eight levels; each level representinga mobile phone) and emotion (ten levels; each levelrepresenting a measured emotion). Figure 10 showsthe two-dimensional solution of this analysis. Thefirst dimension, which is represented by the horizontalaxis, accounts for 69% of the total inertia, and thesecond dimension, which is represented by thevertical axis, accounts for 22% of the total inertia.

The solution in Fig. 10 shows a ‘mobile phone andemotion’ space that visualizes the associations be-tween the phones and reported emotional responses(see also Fig. 2). The distances between the stimulireflect the relationship between these stimuli (withsimilar phones plotted close to each other). Similarly,the distances between the products and the emotions

Fig. 10 Mobile telephoneand emotion space, study 3


reflect the relationship between them. This means thatproducts that are plotted close to each other elicitedsimilar emotions, whereas those that are plotted at adistance from each other elicited different emotions.The space indicates that the new design is one of threemodels that elicit the wow-emotions: pleasant sur-prise, fascination, and desire. Table 4 shows the meanratings of all stimuli on the three wow-emotions (on athree-point scale rating). In addition, the last rowshows the ‘wow-index,’ that is, the overall meanwow-rating.

In Table 4, the models are ordered in accordancewith their wow-impact. The new model in the secondcolumn (model G) has the highest wow-impact,whereas the model in the ninth column (model E)has the lowest wow-impact. Consistent with Fig. 10,the table indicates that models G, B, and D have thehighest wow-impact with the new model (G) as themodel with the highest overall wow-impact. Arepeated-measures ANOVA with the emotions as thedependents indicated a significant difference (p<0.5)between the wow-rating of model G and all othermodels except model B (and between model B and allother models except model G). This indicates thatboth models G and B elicit higher levels of wow-experience than all other stimuli.

Discussion

The aim of this project was to explore how a wow-experience can be conceptualized and how one candesign products that elicit such an experience. It wasfound that with the proposed approach, it is possibleto design a product that evokes a wow-experience.

The approach has three main ingredients. The firstingredient is that it conceptualizes wow as anemotional experience that combines the emotionsfascination, desire, and pleasant surprise. These threeemotions are elicited by particular appraisals that canbe combined to the following ‘wow-appraisal’: theproduct is appraised as unexpected, unfamiliar,promising, and fit for possession.

The second ingredient is that it draws its designparameters from a combination of qualitative andquantitative research in which the wow-impact ofexisting products is measured and discussed withusers to find the relevant concerns underlying thewow-appraisal. The strength of combining qualitativewith quantitative research methods is that it generatesrich and stimulating design parameters, and at thesame time, testable premises.

Third, it focuses on the product concept (in termsof character) instead of on product features as is donein feature-based approaches such as the excitementapproach proposed by Mann (2002). In our view, thecharacter-based approach creates a more durablewow-impact. The experience is not a first impressionexperience (that was measured in the evaluationstudy) only, but one that evolves over first impression,usage, and ownership. The design for wow approachthat focuses on layers fits the conception that productscan evoke multiple ‘layers’ of emotion, and anunderstanding of the various layers that combine toproduct experience enables the designer to create richand desirable products.

At first sight, the concept of wow seems superficialas it is nothing more than a simple verbal exclamationthat comes with excitement. This excitement will lastuntil a new and even more exciting product comes

Table 4 Measured wow-impact in study 3


into view. At second sight however, this experiencesis more complex and rich than a mere excitement. Itis evoked by an intricate and compound appraisaland has an experiential, physical, and behavioralimpact. In that sense, design for wow can result inproducts that people are not only excited about tosee–but also exited about to use and to own;independent of the inevitable future products thatsurpass its wow-impact with the newest up-to-the-minute excitement features.

Acknowledgements This research was funded by KPN. MarcSteen of TNO Telecom, and Christian van Hulten, Pablo Zomand Radboud van Kleef of KPN are acknowledged for theircontribution and support.

References

Arnold, M.B. 1960. Emotion and personality. New York:Colombia University Press.

Desmet, P.M.A. 2003a. A multi-layered model of productemotions. The Design Journal 6(2): 4–13.

Desmet, P.M.A. 2003b. Measuring emotion; development andapplication of an instrument to measure emotionalresponses to products. In M.A. Blythe, A.F. Monk, K.Overbeeke and P.C. Wright (Eds.), Funology: fromusability to enjoyment (pp. 111–123). Dordrecht: Kluwer.

Desmet, P.M.A. and Dijkhuis, E. 2003. A wheelchair can befun: a case of emotion-driven design. In Proceedings of

The International Conference on Designing PleasurableProducts and Interfaces, June 23–26, 2003 Pittsburgh,Pennsylvania, USA. New York: The Association forComputing Machinery.

Desmet, P.M.A. and Hekkert, P. 2002. The basis of productemotions. In Pleasure with products, beyond usability,edited by W. Green and P. Jordan, 60–68. London: Taylor& Francis.

Desmet, P.M.A., Overbeeke, C.J. and Tax, S.J.E.T. 2001.Designing products with added emotional value: develop-ment and application of an approach for research throughdesign. The Design Journal 4(1): 32–47.

Frijda, N.H. 1986. The emotions. Cambridge: CambridgeUniversity Press.

Lazarus, R.S. 1991. Emotion and adaptation. Oxford: OxfordUniversity Press.

Mann, D. (2002). Design for wow; an exciter hypothesis. http://www.triz-journal.com.

Millard, N. 2006. Learning from the ‘wow’ factor – how toengage customers through the design of effective affectivecustomer experiences. BT Technology Journal 24(1): 11–16.

Ortony, A., Clore, G.L. and Collins, A. 1988. The cognitivestructure of emotions. Cambridge: Cambridge UniversityPress.

Reynolds, T.J. 1986. Developing a complete understanding ofthe consumer: Laddering theory, method, analysis, andinterpretation. Oslo, Norway: Institute for ConsumerResearch.

Terninko, J. 1995. Step by step QFD: Customer driven productdesign. Boca Raton, Florida: CRC.

Wakefield K.L., Baker J. (1998). Excitement at the mall:Determinants and effects on shopping response. Journal ofRetailing 74(4): 515–539.


http://www.triz-journal.com

http://www.triz-journal.com

emotional design; application of a research-based design approach

Documents