A prefatory typology of social media design bureaus : the behaviors and characteristic traits of design bureaus involved in social media designing

(1)

A prefatory typology of social media design bureaus – The behaviors and

characteristic traits of design bureaus involved in social media designing.

Name: Carl de Jong Student number: 5950147

Faculty: Economics and business Study: Business Studies

Thesis supervisor: Ton Gruijters

Assessors: Ton Gruijters and Tsui Vinig Date: June, 26, 2015

(2)

Statement of Originality

This document is written by Carl Jan de Jong who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

(3)

Abstract

Social media design is relatively new field of design and design literature does not provide ways to identify the various categories of design bureaus that may exist, nor the corresponding behaviors of such design bureaus. This paper sets out to identify them and ultimately to arrive at a prefatory typology. There is a practical relevance to any stakeholder in a social media platform, including its users, designers and clients. There is also a scientific relevance as there has been no attempt to create such a typology.

The author of this article has interviewed various visual- and interaction designers in four different Dutch social media firms. All at some point have been working on an entirely new social media

platform, yet in very different ways. Video-analysis of online lectures were also used to cross reference and to validate the outcomes of this research.

The outcome of this paper is that design bureaus that occupy themselves with designing social media platforms fall in general into three types. The artistic type, the incrementalistic type and the

commercially driven type. The behaviors of such organizations and the outcomes of their efforts are different too.

The artistic type enjoys great autonomy and creative freedom and their work is relatively open ended, they are idealists, visionaries and tend to use no form of automation whatsoever, they also add

emphasis on elegance and iterate in flux, their work is like a blanc canvas. Incrementalists are more opportune, they set core goals with clients where they cling onto, however other features are created on a more ad hoc basis. They tend to pattern and automate as much (mechanical crafting) as possible and their creativity comes from making adjustments and incorporating into the design, rather than by creating completely new things. They are positivists in that they are both pragmatic and idealistic. The commercially driven type has a more traditional approach. They operate in a bureaucracy where everyone has a clear task according to their specialization. Any effort is a joint effort, they are continuously consulting with each other. They are realists and value quantifiable information over qualitative and functionality over aesthetics. This type of organization is suited for all styles of crafting, possibly due to the division of labour.

This article also discusses the relevance of the existing literature on designing. This research can confirm that in the context of social media design, design practice as described in the literature is valid and applicable at least concerning the aspects that were taken into account for this research.

(5)

1 Introduction

We live in times where social media is everywhere. Since social media have come to exist a lot of companies have tried to do something social. A lot of attempts have been made to create the next cool social media platform, with varying success. Users now have several successful platforms to choose from but how were they created?

There has not been a lot of research that attempts to describe or categorize the design styles of social media-designers. Yet there is relevance for designers, clients of design bureaus and science in general to identify such styles. The design style can have a significant effect on the outcome of the design, it can impact the behavior of a social media platform and anything / anyone connected to it. It is also very relevant to clients to know upfront what sort of design style should be expected, not only in terms of the outcome but also what they can expect during their conjunction / collaboration with the design bureau. Designers may want to know if the design style of some design bureau matches their own way of working or ambition to work a certain way.

Furthermore one could ask if the existing literature about designing and organizational design is applicable to the field of social media. Social media is a relatively new field in which the dynamics are different from many other forms of design; Companies can consist of only a couple of people, can be completely virtual, can rise and decline in value in a extreme rapid matter and users play a greater (active) role than in many other fields of design. Design literature thus far does not suffice to explain the behaviors of social media design bureaus, and this paper can form a starting point in future research.

This paper aims to create a general categorization of social media design bureaus in terms of design styles and corresponding behavior. Ultimately this paper will propose a prefatory typology of design bureaus that are involved with designing social media. This paper looks to find answers from an organizational perspective.

This paper will first establish a theoretical framework and how this paper can further attribute to literature. Next research questions and a methodology will be proposed. After which the result section of this paper will answer the research questions. In the discussion section of this paper the general relevance of this paper in relation to the existing literature, examples that possibly affirm the findings of the result section and the limitations of this research will be brought up. Finally this paper will conclude in the form of a summary.

(6)

2 Theoretical Framework

The practice of design can be seen as a situation in which a designer creates something in his environment through a design process, while at the same (s)he is also part of larger patterns of

designerly ways of knowing, thinking and acting. Simply put the environment helps create, establishes boundaries while the eventual design and the process in turn will become part of this environment be it physical or abstract. Establishing a product for example creates some experience that a designer will remember the next time (s)he creates something similar. The practice becomes part of a larger pattern of designerly ways of knowing, thinking and acting and can become part of entirely different design process in its own right. For example when (parts of) the design is reused in something different. Any design process will involve actors, actions, context and an end-result. A network of these elements are embedded in any design process, meaning that actors will take actions given a certain context and leading to some result. For the purpose of this research 'context' was split up into knowledge and values, seeing that 'experience' is both a form of knowledge and it could change the actor (from an epistemological perspective) other forms of context are hard/impossible to distill from the literature and unaccounted for. Results were also hard to distill from the literature, mostly the object/artifact that was produced by ways of designing in the literature, were not described. This theoretical framework will thus focus on these four elements: Actors, actions, knowledge and values, as parts of a design process.

This theoretical framework will be based on this simple model. And this paper will use schematics of a design process specifying these four elements throughout the theoretical framework in order to clarify similarities and differences between theories about designing.

To clarify these elements, one can specify actors as narrow as client – user – and designer or as broad

as involving every stakeholder involved in the process of designing and using the end product and anything/anyone that has any influence on the design process. Actions are the way designers go about

designing it is what they actually do. Knowledge and values are the context of design process, it is

what they know, understand and what they feel. Any design process aims to have a result, this can be a physical product, an intellectual product or knowledge about designing. These are not exclusive

towards each other, i.e. if one designs a physical or intellectual product one ends op with knowledge about the design process too. This paper will get into the elements of a design process more extensively when discussing ways to look at designing through current design literature.

Existing theories about designing bring up certain aspects that are most typical in any design process. These aspects are most distinctive in the design and this paper will call them key theoretical elements (kte’s). It is what distinguishes one theory from another. This paper will clarify the roles of those elements and elaborate on this model after explaining the dominant literature on designing. Going by one theory at the time this paper will fill in what the kte’s are in the design process with respect to the theory.

(7)

2.1 Traditional ways of looking at designing as practice:

In traditional methods of design we can see how designers have been successful in the past. Jones (1992) proposes that there are two traditional methods, craft evolution (also referred to as design-by-craft), and design-by-drawing.

Crafts

Craft evolution follows a naturalistic way of designing, it is commonly practices by trial-and-error and

craftsmen might not be aware of the reasons for crafting the way they do. Information on the design process goes into the product itself, and could be seen as a sort of ‘genetic code’. Success depends on how well parts of the design can be adjusted relatively to each other, the evolution of complex forms that are derived thereof and how well they are fitted to the circumstances in which they are used. Though this technique may have been useful in the past it is highly doubtful that tinkering with

materials, through trial and error all complex products of todays world can be made. (In retrospect this was a thinking error of the author of this paper, trial and error remains a valid technique even today, as the results of this paper will point out.)

(8)

Design by drawing

Design by drawing has also been influential and many might argue it still is. Design by drawing makes

it possible to ‘split up the production work’, thus allowing for a ‘division of labour’ as Adam Smith probably has meant it. This is because the thinking is separated from the production, thinking happens during the drawings on the drawing board. It also allows for an increase in size (bigger projects), production planning, and an increase in the rate of production (fragmentation of tasks and

specialization). Some of the adaptability of the craftsmanship is lost is this way of designing, while drawing there is no telling if the actual product will be adequate for its circumstances or if there will be a demand for it. Designers thus have to imagine and rely on memory and intelligence to determine future success. Information on the design process goes into drawings, prototypes (and possibly documentation on reflections).

Before looking at designing social media this paper will look at what constitutes successful designing in another context. One of the first fields that may come to mind is probably that of architecture. Architects have been practicing design long before the existence of what we now term ‘social media’. Architects seem particularly fond of the designing by drawing method, though in the current era they may or may not substitute paper with digital drawings on the computer.

Lawson a dean of the faculty of architectural studies at the Sheffield university and author of several books on designing explains that designers, clients, users and legislators all pose constraints on the designing process, the constraints of the first two are considered relatively flexible, while the latter are considered relatively rigid. They can be internal (related to the product itself) or external (related to its context). The purpose of the constraints is ‘to ensure that the designed system or object performs the functions demanded of it as adequately as possible.’ There are an inexhaustible number of different solutions and there is no such thing as an optimal solution to a design problem. (Lawson, B., 2006. pp 83-124) Designing requires a different kind of thinking, this thinking needs to be both convergent (common to scientists) and divergent (common to artists), and successful designers also tend to be both intelligent and creative. (Lawson, B., 2006. pp 153-154) Designers converse with their work. The conversation between a designer and his/her work means that (s)he will make a representation of solution and reflects on it. For example a drawing speaks to its creator and might suggest

(9)

improvements. (Lawson, B., 2004. pp 90)

Designers also posses the ability to allow incomplete and possibly conflicting ideas to coexist and to be at ease with the fact that solutions often come towards the end of the designing process. (Lawson, B., 2006. pp 153-154) ‘This indicates that good designers are able to sustain several “conversations” with their drawings, each with slightly different terms of reference, without worrying that the whole does not yet make sense. This important ability shows a willingness to live with uncertainty, consider alternative and perhaps conflicting notions, defer judgement, and yet eventually almost ruthlessly resolve and hang on to the central idea.’ (Lawson, B., 2006. pp 219) ‘What is common here is the idea that somehow, through a clever mental process, some obstacle or conflict is simply removed by taking a particular view.’ (pp 275) Designers do this by framing, which we can consider a window from which we look at the world. It is likely that experienced designers posses several ways of framing. (pp 275-276). Thus, the designer tries to understand the design problems, identifies and frames them, find ways to represent them (for example drawings), has conversations with those representations (multitude) and then plays with creative ideas and reflects on them.

Since more ways of looking at design have been identified. Jones (1992) however explains that most design theorists agree that in todays world with its complex products we can see design as a three-‐ stage process. First the space for a solution is defined by widening the view on the problem and

extending boundaries as much as possible (divergence). What follows is a phase of seeing and

establishing patterns while identifying critical variables and constraints. The problem can also be split up in multiple sub-problems, making a more feasible solution space (transformation). And finally a

phase in which uncertainties are reduced and an alternative is selected and its (physical) form created

(10)

2.2 Modern ways of looking at designing as practice

However we can ask ourselves how relevant this is in the context of social media. These methods of designing have been largely if not totally based on designing a physical product and social media are only partially physical and this three-stage-process that Jones describes will occur in any method of designing. This paper will therefor consider two more theories of design more related to information systems, human computer interaction (HCI) and design-‐as-‐process.

Human computer interaction

Preece, J., Sharp, H., and Rogers, Y. (2007 pp 418-453) explain that in HCI designers identify requirements, design a solution that satisfies those requirements and evaluate. In this process it is fundamentally important to generate lots of alternative ideas, involve users and other stakeholders in the design process and communicate lots with people other than the designers through means of sketches, descriptions, and prototypes. They propose a user-centered-approach, meaning an ‘early focus on users and tasks’ (through observation and trying to understand user-behavior), ‘means of empirical measurement’ (identifying goals with utmost specificity), ‘and iterative design’ (trial and error, feedback-loops) as means of being successful at designing in the HCI-field.

As ways of iterative designing one can use scenarios and prototypes to get feedback from either

speculating or from the (potential) users. In case of scenarios the designer or designer-team has rely on the intelligence that goes into this scenario-making. In case of prototyping the designer(s) can emulate the circumstances of the product by exposing it to (potential) users and observing. (pp 554-566) Designers than have to combine three methods of evaluating, ‘usability testing’, ‘field studies’, and ‘analytical evaluation’ through means of observing and asking users and experts, testing users’ performance, and modeling users’ task performance. This would allow to predict the efficacy of the design. (pp 591-595)

We can see that this way of designing has a lot in common with both craft evolution and design by drawing, though neither a physical product nor a drawing has to be made per se.

(11)

Though this theory is likely to explain why some software companies are successful and others or not, it is not sure that it is completely applicable to the concept of social media. It emphasizes greatly on the user experience in terms of ease of use, and the efficacy of the user interface. However social media like Facebook and Twitter might not be that user friendly and still be popular. Hart et al (2008) found that Facebook at the time of writing their article performed poorly in respect to traditional usability guidelines dominant in HCI. They suggest that the positive user experiences gained from using

Facebook might outweigh the rather poor usability. ‘As a web service it not only provides a great deal of social pleasure but provokes curiosity, provides a base for self expression and evokes memories of the past, along with a myriad of emotional and hedonic user experience.’

Design as process

Design as process is designing by singling out and objectively studying them in a positivistic manner.

Design is goal-orientated and design activities are aimed at creating an artifact or system. Steps of designing this way are to define the problem, identify the field of solutions, implementation of the chosen solution and evaluating the implementation. Issues of system functionality and design are central and implementation success is measured and evaluated in terms of utility of the system. (Maris, I. (unpublished)) There are some similarities with HCI, however HCI evaluates on user experience while design as process focusses on utility.

This way of designing might work if the problem is identifiable, and the designer(s) can see and identify the field of solutions, however this might not always be the case. Measuring in terms of utility is trying to rationalize something subjective. Designers might be able to benchmark but only against their own preconceptions. And some designers might come to solutions intuitively rather than to

(12)

consciously choose a solution. It also focusses on utility, while as for the critique on HCI the popularity

of a particular social medium might be very much based on the emotions it evokes.

Thus the before mentioned theories are all lacking in some aspects to explain what makes a social media design strong yet they may identify (partially) what kte’s really matter. They are helpful if one tries to explain what makes a design in general strong as is but they have their limitations. While crafts-evolution has been around the longest and can explain pretty much every design before the twentieth century, it would fail dramatically if we use it to analyze the emergence of social media because the object or artifacts needs to fit the needs of a multitude of users and that requires input of those users. Design by drawing has been very influential until very recent times and in some practices it still is (for example arts, fashion or architecture) but it also fails to account for the role that for example users can play in a design process. HCI does focus on the users and other stakeholders, however the ways of evaluating seem to be forcibly objective and tends to loose sight of any subjective factors that matter, such as the emotions a design evokes as has been explained by Hart et al (2008) in respect to the popularity of Facebook.

Practice theory and actor-network theory provide a fresh perspective. These theories are applicable in social studies and particularly helpful in ways to look to information systems. This paper will first explain these theories by means of a quick review, because a simple explanation of these theories will probably not suffice to give a good overview. Both theories are a tad complex and a bit philosophical. Thus I briefly summarize some of the findings of influential authors before going into how they provide a fresh new perspective and their application in the context of designing social media.

Practice Theory

Practice is a way in which designers order their surroundings into a structure that makes sense to them. It is ordering heterogeneous items, implicating that objects can take on various forms, be it physical or

abstract. (Gherardi, S. 2009)

Objects are meaning-producing and practice-generating; objects both signal what can be done with them as well as their lacks. Even when an object is a product and considered ‘complete’ it still signals room for improvement. This is especially apparent for epistemic objects where the defining characteristic is its “changing and unfolding character - or its lack of ‘object-ivity’ and completeness of being, and its non identity with itself [..] epistemic objects have to be seen as transient, internally complex, signifying entities that allow for and structure the continuation of the sequence through the signs they give off of their lacks and needs” (Schatzki et al. 2001)

Designers produce and reproduce socially sustained patterns of action, which may take on the form of an physical or abstract objects. Attachment to an object of practice makes it that practices are socially sustained by (communicated) judgments (i.e. gestures, tacit negotiation, appraisal) related to ethics and aesthetics. These patterns of action define taste as much as they are defined by taste in a constant negotiation in a community of practitioners. Practices are thus constantly (subjectively) refined through a taste-making process striving for perfection, never achieving it. Taste is learned, it is

determined by the social surroundings of an individual or group.

“We may see taste as ‘a sense of what is aesthetically fitting within a community of

practitioners – a preference for the way we do things together. [..] Taste is learned and taught as part of becoming a practitioner and it is performed as a collective, situated activity – taste-‐making – within a practice.” (Gherardi, S. 2009)

(13)

Helen Kennedy found that what professionalism constitutes in the context of a community of web-designers is very similar; it means a negotiation between the designer and the community about what is good or bad practice in a self-regulating community that is open to participants and critical to the participants’ work. (Kennedy, H. 2010)

Designers should also be aware of the communication of their product/artifact. There is (almost) always a discrepancy between the designers’ intention and the users’ interpretation. Inevitably it seems to happen that users try to infer what the intention of the designer is, by reasoning and assuming. These

inferences can assist users in determining what a product is, what its use is, and how it should be

regarded. Ultimately it may even determine its meaning. (Crilly, N. 2008)

When a stable way of doing things is institutionalized and made normatively accountable for both the practitioners and outside viewers it becomes a practice. To empirically analyze practices, Gherardi proposes three levels of analysis; objectivism (practice from outside), subjectivism (practice

from inside), and practice as social practice (or what implications it will have on the larger society).

(Gherardi, S. 2010)

The nature of design in a sociotechnical system requires knowledgeable contributions of

designers in both social- and technical elements. The tools designers use are closely intertwined and co-evolving and it is uneasy to see where the choices they make will lead them, upfront. These tools have technical and social affordances; they determine how easy (if possible) some actions are, while they

also determine the strength of support for it. (O’Day et al. 1998)

The term affordance is used to describe an attribute of an object that allows people (designers) to understand how to use it. (Preece et al., 2007)

Wanda Orlikowski and Susan Scot have analyzed existing literature concerning technology in organizations. They found in general two streams, that of discrete entities and mutually dependent ensembles, which respectively focussed too much on the technology (techno-centric perspective) or the social (human-centered perspective) within an organization. In fact assuming that technology and human actors are different and separate entities may blind us from seeing how material (technology) can be an integral part of organizational practice. They propose a more adequate third stream

“sociomaterial assemblages” as an umbrella for existing concepts in the literature that do not separate humans from technology but rather study their relationship.

“.. entities (whether humans or technologies) have no inherent properties, but acquire form, attributes, and capabilities through their interpenetration.”

Relationships and boundaries are enacted in practice. Thus “practices are always sociomaterial, and this sociomateriallity is integral, inherent, and constitutive, shaping the contours and possibilities of everyday organizing.” (Orlikowski, W.J., 2007 ; Orlikowski, W.J., and Scott, S.V., 2008 ; Orlikowski, W.J., 2009)

In case synthetic worlds (such as social media) are to be studied adopting a sociomaterial perspective seems particularly useful as it highlights how this synthetic world is not neutral or determinate but material and integrally a part of the phenomenon. (Orlikowski, W.J., 2009)

From practice theory we can derive a lens to look at the phenomenon ‘social media’, looking through a practice lens. The practice lens is a way that allows studying people in their interactions with

technology through practices that enact structures which shape their emergent and situated use of that

(14)

Looking at the design practice from a practice theory lens designers should keep in mind that, • Objects in a social medium (network) are both meaning-producing and practice-generating. • Structures emerge through practice and are sustained by it, a social medium can only exist by

continuous practice.

• Practices are socially sustained by judgements related to ethics and aesthetics and practice is always sociomaterial, a social medium must therefor be both social and material.

• Taste-making refines practice, taste is both learned and taught by practitioners, it is a collective situated activity in a (self-regulated) community, taste-making should be an integral part of any social medium.

• Inferences help users determine the meaning of objects/artifacts through reasoning and assuming, a social medium or part of it is subjected to inferences and therefor this determines its meaning.

(15)

Actor-network theory lens

“I plus the car, plus the dozens of patented engineers, plus the police are making me be moral.” Prescription is the process where a non-human (artifact) imposes moral on a user through its

moral and ethical dimensions. In a actor-network efforts can be delegated either to humans or non-humans. Delegating the opening and closing of a door to non humans (hinges, motors, motion detection, hydraulics) is given as an example in Latours’ article. This delegation transforms an effort into a smaller one, this process is called translation (among other synonyms). There is a trade-off in

this process, unskilled (untrained) humans require skilled (smart) non-human actors (artifacts), and vice versa. (Bijker, W. E., & Law, J., 1992)

Humans script the way an artifact has to behave, while as a result inevitably the non-human will also prescribe human behavior. As is the case with the door example, if motion sensors and a opening-closing mechanism are applied to a door, the human has no choice but slow down his/her walking when approaching the door and waiting for it to open.

Any complicated mechanism is filled with contradictory specifications, they are inscribed with them. Ideally one would walk through the doors, without having to wait or slow down and still keep the warmth in/out, while safety issues also dictate that the doors should not open or close too fast possibly causing injuries to occur. “The program of action is in the answer to an antiprogram against which

the mechanism braces itself.” An artifact is only part of a program of action and of the fight against antiprograms. Faced with programs of action one can only observe “how a negotiation to associate dissident elements require more and more elements to be tied together and more and more shifts to other matters.” (Bijker, W. E., & Law, J., 1992)

Society itself can be seen as a heterogeneous network, consisting of both human and non-humans. Latour argues that society is constructed but not socially because we exchange properties with

non-humans through technical delegation. “Humans, for millions of years, have extended their social relations to other actants with which, with whom, they have swapped many properties, and with which, with whom, the form collectives. [..] Responsibility for action must be shared, symmetry restored, and

humanity redescribed: not as the sole transcendent cause, but as the mediating mediator.” (Latour B., 1993)

A social network is formed through the collective of human beings and endless other materials (actors). If the materials were to be taken out of these collectives, the social order would cease to exist. In actor-network theory “there is no reason to assume a priori, that either objects or people in general

determine the character of social change or stability. [..] an actor is a patterned network of

heterogeneous relations, or an effect produced by such a network.” Hence an actor is also always a network, and all networks participate in the social. Social structure is a verb rather than a noun, it is a relational effect that both generates and reproduces itself. (Law, J., 1992)

As a network grows/extends it intensifies network (inter-)dependencies. As this happens human actors might try to counter this complexity, trying to tame technology, working to make it more

efficient and predictable. But in doing so they also program their own behavior (if an artifacts

flexibility is decreased, then the flexibility of its human counterpart is also decreased). With respect to an online learning community, Thomson found that attending to the mix of actants, its configuration and the level of attention actors are given can help make informal learning practices distinctive. (Thompson, T.L., 2012)

The actor-network theory lens should enable us to look at a network and observe relationships with the various heterogeneous bits and pieces that constitute this network and study the roles of actors in it, be it humans or non-humans, and how this network is juxtaposed, sustained or disbanded.

(16)

Looking at the design practice from a actor-network theory lens designers should keep in mind that, • Translation is more effective if it anticipates the responses and reactions of the materials to be

translated. Appropriate relational circumstances have calculational consequences, which increase network robustness of a social medium. (If the design anticipates inferences of users it will more likely be used the way it was intended.)

• We may see a social medium as a product of a series of strategies which operate to generate complex configurations of netwerk durability, spatial mobility, systems of representation, and calculability. (For a social medium to be successful choice have to be made, all choices combined can be seen as a complex configuration of choices on which the likelihood of the favor of its users will depend.)

• Any actor in a social medium (be it physical of abstract) influences the entire social medium (network).

• Any social medium is expected to have some contradictory specifications and programs of action that overcome antiprograms. (Overcoming one problem sets limits or poses a problem for something else. (Like the automatic door example. A design is always a

configuration-compromise)

• The size of the network (social medium) is expected to have considerable impact on network interdependencies, meaning that the larger the network becomes the more it will depend on its actors and in turn their dependencies. (For example Facebook started in the US, when it went overseas the servers may have had to be adjusted or may have bought new ones because in a 24 hour a day environment there can be no downtime for the servers for maintenance etcetera. Also simply tagging common in Facebook use, may had added severe complications to its relational databases because the friends of friends-relationship should now be actively shown on peoples pages. X sees that Y was tagged in the album of Z and so on. And the advertising income should depend on how many people are online at that particular moment etcetera.)

(17)

To recap the before mentioned sociomaterial assemblages. While practice theory focuses on the way designers order their surroundings that consist of heterogeneous items be it physical or abstract into a structure. During this process of designing objects both signal what can be done with them as well as what they still lack. When designers design they use patterns of action by which their way of doing things is defined (taste) but they also define these patterns themselves (taste-making). Actor-Network

theory in turn focuses on the continuous interexchangeability of actors within a network without

valuing one more over the other. Humans exchange properties with materials/objects and vice versa, humans define the network as much as they are defined by it. To point out one of the most obvious examples, todays humans at least in the western civilization could not provide food for themselves if they had to hunt, rather they need the convenience of a butcher shop or supermarket to eat meats, thus the network (society) defines how people live, while if all people decide to abandon this way of living they would probably become able hunters (again) or find another solution.

The theories of practice theory and actor-network theory are more adapt to account for matters in a design process that the other theories do not. It transcends the traditional views of designing in that it holds entire communities or networks of actors responsible for designing. While in the traditional views the designing-role always lies with the designer. Emphasizes was added because the traditional theories always hold one person or a limited group of people responsible for the design. In the world of social media however it is not unlikely that the medium itself adapts to the input of large groups of people be it regular users or otherwise, like the hashtag-use with twitter, which was not part of the design nor was it planned to become, it just emerged.

The KTE's of this theoretical framework will be used in the discussion part of this paper to explain the relation that exists between the literature i.e. what it is that designers do according to the literature (KTE) and what designers do in practice (reality). This paper will establish that these design-theories and derivations thereof are inadequate to identify the different types of design bureaus of social media platforms that exist in the Netherlands. However this paper will also show that these theories are still very relevant in explaining what designers do in their everyday work.

Thus the design-theories of this theoretical framework explain what it is that designers do when designing, it does however not explain what type of design bureaus exist and what characterizes bureaus that design social media. What sets one apart from the other if they are all involved in designing social media platforms? Research towards this characterization is needed. A prefatory typology is needed to explain the different behaviors and characteristic traits between the social media design bureaus that exist.

(18)

3 Research questions and methodology

This article aims to come to a prefatory typology of social media design bureaus.

1. What types of typical behavior can be identified looking at social media design bureaus? 2. What character traits do those types of behavior reflect?

3. Accounting for these character traits, what type of design bureau style does that resemble? Interviews

Interviews were held at 3 social media designing companies/bureaus with both interaction designers and visual designers (Matte, Fabrique and Unitid). There was also a fourth company (Hyves) were only the creative director was interviewed. Most design teams within these companies composed of these type of designers sometimes they were assisted by (external) programmers and/or client-experts. But at the very core of a design team there were always a interaction designer and a visual designer. Thus they were the focus of this research. Another criterium was that all the participating companies had to be involved in designing a social media platform themselves, hence not just additions to an existing platform. The fourth company will be brought up in the discussion. The interviews conducted took on average one hour, where 28 questions of semi-open nature where asked. The questions were directory rather than binding and interviewees were invited to respond in an as open as possible way. The author of this paper also tried to not use jargon or terms from the literature to solicit the most open and

spontaneous answers possible. To rule out socially- and professionally desirable answers as much as possible and to get honest answers about the behaviors of designers. The list of questions can be found in the appendix (1).

The analysis of the interviews

These interviews were transcribed and then put into Atlas Ti (software for qualitative research). From there the transcripts were coded openly. (Phase 1) This code-list was put into Excel software to count the number of times they were mentioned as a starting point to see the significance of those codes. Also by comparison the author of this paper was able to translate and further the meaning of the codes, while memos and notes could easily be added to the Excel files. The transcripts were then further analyzed and recoded in Excel in a cyclical fashion to further the meaning of these open codes. Eventually meaningful axial codes could be distilled. (Phase 2) These codes were in turn used to come to a

description (selective coding) of the companies that were visited and to compare the companies to each other in terms of characteristics. (Phase 3) Full descriptions are included in the appendix (2). The sum of the characteristics and especially the differences between the companies have led this paper to very clear distinctions as well as being able to keep track of similarities.

From there a list was devised in which a comparison was made between the companies in terms of design according to their general description and the code lists. A visual representations (by example of) this list can be found in the appendix (3). In this list there were 9 main categories found, by which the companies differentiate themselves as well as some traits that are distinctive but unique. (Phase 4) Scales were created to visually represent these distinctive categories. These categories were aligned in order to see what kind of organizations the design bureaus really were. An alignment of three

categories at a time was made, to ultimately derive at a conceptual model of a typology for social media design-bureaus in the Netherlands.

(19)

Phases of the analysis explained

Phase 1: Open coding, in this phase all data that has been collected has to be divided into fragments. After which the fragments need to be compared among each other and grouped into categories where they come to mean the same subject. This is done in a cyclical fashion. Eventually these codes will come to mean a summary of certain segments of the data which will make interpreting the data easier for research, as it enable easy retrieval of fragments that have been assigned a specific code. (Boeije, 2010, p. 96-98).

Table 1.1: Examples of open codes Open code Transcript Asking for

help

You can just give it to the programmer and say 'can you make this for me ?'. And then I'll contact the programmer.

Sometimes ask people what they think of it.

I ask colleagues or friends or anyone with a fresh-eye. I guess you try to work with other people.

..And then he would sit with the developer and just try to explain things.

..And I'd give some kind of description... And the developer would create something for it. Create a

strategy to solve a problem

To create a strategy

What you want to say and how you're going to. And how to use tools to apply what you want to say.

Design to me is to create a strategy. You have to something that you want to bring to the world, or to other people.

I'm always looking at what is needed, so you try to find a solution for it. Then I always think, 'how do I do it?'

How it could look and how it should feel, also in terms of non-look. Then I'd think up a concept, just think it up, just create sketches.

Not just getting started with the project but to set milestones like: 'Ok, we'll be doing this then and then.'.

To figure out what the goal is, and what the thing is you're trying to achieve and to think of a concept for it.

We often have dilemmas and people have to be able to deal with them, for example by creating a number of alternatives and to test them.

Looking for alternative solutions

..but being able to find certain solutions.

But also to see what the alternatives are. To look if you can do something better, or that there is another way to do it.

So I guess because the interface is rather limited, you find other ways to make it interesting. I think to me there doesn't need to be just one solution.

..may know a better, smarter solution.

Ask yourself how can you make it even better?

And what you can add to it. Or maybe this problem brings up more questions and you can't solve this problem. So you ask more questions, that would also be fine.

You always have to be quite inquisitive in finding different things.

You have to find an alternative solution.. I have no idea what, but there must be a solution. If you find that something is not working, try to find a work-around.

..just look at the problem first like 'ok, something else might be more interesting for this.' To achieve the same goal, but in a different way.

..we try to redefine the problem, so we can come to different solutions too.

Perhaps it are the moments where a pattern is disrupted where you come to enrichment. Phase 2: Axial / focussed coding, in this phase all the open codes are reassembled into categories (axes) and sub-categories. It also specifies properties and dimensions of a category. The fragmented open

(20)

codes are now put together into new data, thus it produces new meaning and moves in the opposite direction of open codes (codes to data instead of data to codes). This also happens in a cyclical fashion as the relationships between salient categories and sub-categories can always be modified, refined, elaborated upon and rejected during this phase. The primary purpose of axial coding is to make clear distinctions between dominant elements in the research and the less important ones. Secondary it reduces and reorganizes data sets (check for redundancy and synonymity, while selecting the most important data). Eventually axial codes will be able to explain larger parts of the data and reconnecting them, while codes that do not fit are forgotten (unless researchers explicitly make attempts to not forget e.g. create memos). When new data is presented and there is no longer a need for new axial codes, researchers can move on to the next phase. (Boeije, 2010, p. 108-114)

Table 1.2: Examples of axial coding Axial code Open codes

Steering • Designing for social interaction, manipulating how people go through the system.

• The outcome is flexible but the goals are set.

• Steering how people use things. • Satisfaction from steering. Creating an

infrastructure

• Design, look and feel in terms of visuals, navigation and structure.

• Structure first visuals later.

• Making structures visible. • Designing; structuring concepts. Real life to digital

interactions

• Translating: Real life interactions to digital interactions. • Digital world catching up to offline-world in terms of interaction. Vicinity testing

and feedback

• Vicinity testing and feedback.

• Finding someone that can provide a helicopter-view. • Asking someone 'blanc' / with a 'fresh eye'.

Understanding the audience

• Fundamental human behavior.

• Psychology of user. • Personality of users. Looking for

alternatives

• Looking for alternative solutions.

• Finding another way. • Seeing other options. Solving problems

by puzzling

• Solving problems by 'puzzling'.

• Problem solving.

• Satisfaction from overcoming circumstances / limitations. Automation • Assimilation / automation.

• Developing patterns.

• Tools to create patterns more quickly. Modifying

elements and patterns

• Modifying (look and feel).

• Simplifying • Tweaking. Users get it • Users 'get it'.

• The design should be self-explanatory.

Phase 3: Selective coding refers to looking for connections between axial codes (i.e. categories and sub-categories) to make sense of what happens in the field. The axial codes have to be systematically

(21)

connected to each other, further refining and developing meaning. The so-called core category describes and explains the researcher's observations. It has to be central and linked to a lot, it has to account for a large portion of the variation in the behavior of data, has to appear frequently, is not easily saturated as category, can be formulated in a more abstract way and helps to facilitate the analysis. (Boeije, 2010, p. 114-118)

Table 1.3: Example of selective coding Selective code Axial codes

Predefinition of end-result (scale was made from

'blanc canvas' to 'framed')

Steering

Creating boundaries Creating an infrastructure Stimulating behavior Abstract to concrete

Real life to digital interactions

Complex solutions to understandable tools

Not an artistic expression rather providing a service Blanc canvas effect

Platform goes its own way

The intentions of the designer are clear to the users

Phase 4: Creating a conceptual model, from the selective codes this paper has distilled overlapping integral categories. In total there were 9 selective codes to explain the larger majority of the data. Integrative procedures can be used to extract interesting elements from the data and to make further analysis easier. Boeije (2010), describes ten heuristics for discovery from which six methods were used, namely to create visual displays, creating a matrix, reading memos, searching and counting, looking for the core category and formulating a typology. Four can be identified in the results and methodology. 'Visual displays' have been used in the form of slider-scales (presented in the results), in the process of coding from open codes to selective codes the heuristic of 'looking at the core category' has been used (presented in the methodology), 'memos' were used to further provide context, ultimately this paper was able to distill a 'typology' (presented in the results). Two other heuristics were used which will be exemplified in the appendix (4). Namely the use of matrices as well as searching and counting.

This paper has further tried to derive meaning from the slider-scales by looking at what it visually represents. The slider-scales both shows the differences and the similarities between the companies in the data sample. The author of this paper then reorganized until it made sense and could be confirmed by the transcripts. This paper has come to three different types of behavior for social media design-bureaus, which will be brought up in the results section of this paper.

(22)

4 Results

The results section will provide answers for the three research questions this paper has posed in the previous chapter per question. This leads to nine characteristic traits that design bureaus show. These traits are in turn used to identify three types of bureaus. Then a general description of these types is presented in which the characteristic traits are explained.

Ad research question 1: What types of typical behavior can be identified looking at social media design bureaus?

In total this paper will bring up nine distinctive categories of behavior by which companies are measured against each other. The categories have been visualized in the form of slider-scales. To see how these categories came to exist, this paper refers to the methodology and to see an overview of these categories in the form of a slider-scale (as a heuristic), this paper refers to appendix(3) of this paper.

Categories:

• Pre-definitions of end-result • Transferability of solutions

• Generalistic vs specialized personnel • Use of patterns / automation

• Design procedural focus • Consultation per interval • Personality of designer • Research methods

(23)

Ad research question 2: What character traits do those types of behavior reflect?

After carefully weighing the companies and filling in the slider-scales, this paper has made an attempt to find meaning by re-ordering these categories of behavior so that they point out to the same kind of character traits.

Three categories of behavior were found and merged this way: • Pre-definitions of end-result

• Design procedural focus • Research methods

These three categories point out to the behavior of designers to either go in a project head first, do things on the fly not knowing where the project will lead them versus designers that carefully plan out everything from start to finish and are able to clearly point out the outcome.

(24)

Three other categories of behavior that could be found and merged:

• Transferability of solutions

• Generalistic vs specialized personnel • Use of patterns / automation

These three categories point out to the behavior of designers to either start over with a project and really save nothing e.g. no patterns, no templates, no forms of automation for a next project versus a company that tries not to reinvent the wheel and re-use as much as possible.

(25)

Finally yet three other categories of behavior that could be found and merged: • Consultation per interval

• Personality of designer

• User appreciation (what matters most)

These three categories point out to the behavior of designers to either think about a project in a certain way. If the project is dreamy, abstract, idealistic, the designer is allowed a great deal of freedom, in which (s)he can create a subjectively tasteful product that pleases the designer very much versus designers that operate in a very systematic and scientific way, where there is little room for personal preference and taste and freedom is very limited.

Figure 1.3:

Three characteristic traits can be identified from these mergers of categories per merger. This presents 9 characteristic traits in total, the table below will show these traits.

Table 2.1

Style of planning Adhocracy Opportunistic Bureaucracy

Style of working Daedel crafts Artisan crafts Mechanically

(26)

Ad research question 3: Accounting for these character traits, what type of design bureau style does that resemble?

This paper will propose three types of social-media design bureaus according to the previously mentioned characteristic traits:

The artistic type of social-media design bureaus.

An artistic type of social-media design bureau operates in an environment of adhocracy, its work can be qualified either as daedal crafts or artisan crafts and in terms of the employees they can be

considered visionary.

The incrementalistic type of social-media design bureaus.

An incrementalistic type of social-media design bureau operates in an environment where opportunity is recognized to set and follow goals, its work can be qualified either as artisan crafts or mechanical work and in terms of the employees they can be seen as positivists.

The commercially driven type of social-media design bureaus.

A commercially driven type of social-media design bureau operates in an environment or bureaucracy, its work can be qualified either as daedel crafts, artisan crafts or mechanical work and in terms of the employees they can be seen as realists.

Schematically depicted: Table 2.2 Type of design-bureau: Adhocracy, Opportunistic, Bureaucracy Daedel crafting, Artisan crafting, Mechanical Visionary, Positivistic, Realistic Artistic type

Adhocracy Daedel crafting or _{Artisan crafting} Visionary Incrementalistic

type Opportunistic

Artisan crafting or

Mechanical Positivistic Commercially

driven type Bureaucracy Any Realistic

Company-specific information about the visited companies can be found in the appendix, in the next section this paper will present the generalized result of the design bureau types.

(27)

The artistic type of social-media design bureaus

Adhocracy

In terms of predefining an end-result, this does not happen. Rather designers try to keep open ends as much as possible, meaning that the social-media platform can always change and is not very framed. If an analogy were to be drawn, think of a painter that does not know how his/her painting will work, yet (s)he does have some ideas on how to paint. Perhaps it resemblance to how Bob Ross used to paint in front of TV-cameras in the 'Joy of painting'. Though it is not completely like it because the designers in an artistic type organization also have to take in feedback from users and the client on the fly and work with it. What is also typical is that they perceive the end-result of their work as being a blanc canvas, they believe it is actually the users that create a picture. So they have really no idea were the project will lead them from start. This also means that their way of working is not highly goal-orientated, rather it happens on the fly and in flux. The designers are given a lot of freedom for their creativity and they can focus, refocus or change focus of the design procedures as they see fit. For example a designer might find that (s)he knows too little about a subject and study it or ask for help or look for authorities on the subject and interview them, having little to worry about procedure and thus (s)he may delay the project temporarily. Finally designers also tend to use qualitative methods for their research more than quantitative, because they are in general not aiming for speed optimization, efficiency and so on, rather they aim to come to the most beautiful and elegant solution possible.

Daedal crafts or Artisan crafts

In terms of the transferability of solutions, designers in artistic type of social-media design bureaus tend to not even try to reuse solutions. It is beneath them. They enjoy their artistic freedom and their challenges to come up with something creative too much to standardize in any way. Of course in an epistemological sense this does happen however it is limited to just that. In the company that was visited any patterning or forms of automation were disregarded. However it could also be the case that a certain base of the project is patterned, for example templates for wireframes or something similar. An analogy could in that case be drawn to a furniture maker, that has wood sawn to a certain size for the base of a table, leaving slightly less creative freedom but still a lot. This would be the difference between daedal- or artisan crafts. Designers tend to be more specialists than generalists. This is because they want to excel in what they do best and thus focus on what they do best and leave other things to co-workers. A good analogy would be that of a musical artist that may actually be able to play four instruments but rather plays one as a part of an orchestra. This also means that (s)he does not need to memorize all the notes, the same applies here, a visual designer for example does not need to know much about programming, though it may be useful in both cases.

Visionary

Designers in an artistic type of social-media design bureau do not have to consult with colleagues much, they do what is asked of them and then leave the rest to others yet remain standby to resume when needed. With the client it is a different story, because the enjoy so much freedom it is mandatory that they inform the client from time to time, showing them what they have been doing and getting re-endorsed every time. If not, the project could come to lead its own life potentially disappointing the client. So periodically though not very often there is consultation with the client and colleagues. The personality of the designer is very idealistic, the designer must be able to fully empathize with the users because they are generally creating something completely new, these designs have not proven

(28)

on that. Because designers are so idealistic, the author of this paper found during the interviews that they also reflect on the world around them in a more philosophical and abstract way than others designers. What matters most to them are the users, and in terms of elegance versus functionality they would favor elegance, because this is what makes them creative and defines them as designers.

The incrementalistic type of social-media design bureaus

Opportunism

Social-media design bureaus of the incrementalistic type neither completely predefine the end-results of a project nor do they leave it all to ad hoc decision making. Rather they start with a goal in mind, something they aim to achieve, to point them in a direction. In the company visited they termed it a 'core goal' from which they built. This opportunism leans slightly more towards having it framed, simply because the fundaments of a project will not change on the fly. Design procedures are more in flux than very goal-orientated, because once the 'core goal' has been set, the company becomes very flexible in how they work. Designers will come up with new ideas all the time and try to fit those things in in the project. The teams are actually not all that rigidly formed either, if someone feels (s)he has something (s)he can add to project, it just happens. In terms of doing research designers lean more towards qualitative research than to quantitative because when designing in this environment of opportunism, adding new features etcetera, on the fly, it may be hard to pose questions that can be answered in numbers. For example were some designer to borrow a pattern from another project or even from another company, what would that come to mean? It's hard to hypothesize an increase in the number of users, while it is easy to figure out how it makes them feel. Of course quantitative methods are not completely ruled out.

Mechanical crafting

Designers that design in a incrementalistic way are very keen on patterning what is good. So they do not force themselves to forget everything they have learned but try to fully exploit there work. This means learning goes beyond epistemology, work is kept, along with notes for future reference. This can either be shared with the world through a database or the company can keep it to itself. However in these times it is more likely that they share as it will increase their popularity and whenever something becomes a standard, companies that do not follow often loose popularity with their users. For example the timeline and the way new stories are on top of the page is now a standard. Would a company now deviate they should have very good reason for doing so or probably face being rejected. Personnel tends to more generalistic than specialistic because of the composition of teams, teams are really not rigidly formed and because of the nature of the increments, designers think of something and

immediately propose it to other team members. They share this work load so everyone has to be able to help out in the area of someone else' expertise. This is not to say that there is no specialism at all, designers do specialize in the sense that they focus on certain solutions, patterns or even platforms. For example designers could all be able to program interaction for mobile apps, yet one specializes in Android, another one in Windows-mobile and yet another one for iOS. Visual designers could

specialize in parts of the Adobe software library, autocad, video-software and so on. At the end of the day they all know enough to help each other out at least.

(29)

Positivism

Designers have consultation with the client with a certain interval, consultation with colleagues is almost continuous. This is because of the ad hoc nature in which designers share what they have thought up with the coworkers. Using many ways to communicate with each other by mobile apps, traditional chat-applications, e-mail, social media, in the cloud, walking over and showing, screenshots and so on. There is a great deal of communicating. In terms of the personalitiy of designers, they are in general between idealistic and pragmatic. Because they do keep an open mind and always try to come up with things that can benefit the project even during design, they have that in common with the artistic type of design bureaus. Yet they are also pragmatic, designers will not stall the project just because they want things to be more elegant, when a decision to implement is made, it is made so. They also keep holding on to their core goal much, thus the fundaments of the project will not change on the fly. This is also the case in terms of elegance vs functionality, designers will decide on core functions that have to work without hiccups and then figure out how to make it look good and possibly new features that add value to the platform. Never the other way around.

The commercially driven type of social-media design bureaus

Bureaucracy

Projects within the commercially driver type of social-media design bureaus are highly framed. This means upfront there is a decision on what project to make and designers stick by it. Alterations and changes are made in service of making the predefined project better. For example user tests may show that the project needs reconfiguration in the sense that buttons needs to be reorganized or an app should load faster and so on, but fundamentally the user tests will not impose a rigorous redesign of the

project. Designers work very systematic and divide the work among themselves according to specialism so top efficiency can be achieved. These types of bureaus have a preference towards quantitative research methods because they are interested mostly in numbers. Things that matter most are expressed in numbers, how efficient an app works, the tax on a database, how many users install the app or use the website, the learning curve, et cetera.

Daedal crafts, Artisan crafts or Mechanical crafting

Where the interviews were conducted the company' employees showed all elements of being of the daedal crafting type of designers. However the author of this article sees no reason why it could not be any of the other types. All the elements with the exception of specialism vs generalism can vary for this kind of organization. If solutions were to be transferable this would benefit efficiency. On the other hand because preceding the design process a specific decision has to be made and designers follow this rigidly, it can sometimes rule out the use of (universal) patterns and solutions. Though designers do tend to be specialists rather than generalists, because teams are composed this way. Companies that work this way believe that efficiency comes from doing what one does best, while at the same time it ensures the highest quality of work.

Realism

Designers continuously or very close to continuously communicate with each other about what they are doing. In fact the company that was interviewed said to never work alone and one designer even said that working alone would be the most stupid and unproductive thing one can do in this line of work. So once a design team is composed they are in contact with each other until the project is finished. This also makes that designers have to have a pragmatic personality, decisions have to be made because

(30)

un-decisiveness stalls the project. Designers also have to be more outspoken and able to absorb critique than with the other type of design bureaus. If designer A coming from his/her expertise thinks that what designer B is proposing either is impossible or unachievable at the moment or just not smart, (s)he will express this. What follows is a discussion and at the end of this discussion there has to be a decision, there is no room for stalling and the team members also have to be able to look each other in the eye afterwards. In terms of elegance versus functionality, designers very much favor functionality, they believe a product should just be very good, beauty is subjective. Of course they do not want to deliver tastefulness projects and try to make it look really good, it does not become a goal on its own.

A prefatory typology of social media design bureaus : the behaviors and characteristic traits of design bureaus involved in social media designing