Nov 2 2016
The architecture of cross-channel ecosystems — From convergence to experience was written with Flávia Lacerda and presented at the 8th International Conference on Management of Digital EcoSystems (ACM MEDES'16), November 2 2016, Hendaye, France. This is a camera-ready draft. The original paper is available at the ACM Digital Library.
Convergence has been quietly reshaping not just media consumption but also the way we interact with products and services. Read/write access through personal and ambient devices enables the creation of free-flowing, actor-defined experiences that connect physical and digital artifacts, people, and locations into information-based ecosystems that vastly exceed the boundaries of the media industry. All sorts of everyday activities, from traveling to education to healthcare, are affected. This paper details the ways convergence is conceptualized in media studies and within product and service realization; argues that while media studies offer more mature descriptive frameworks, design practices are ill served by them as they are descriptive and not generative frameworks; proposes a formulation of cross-channel experiences as an information-based design artifact and of cross-channel ecosystems design as a pragmatical, actionable approach for dealing with convergent experiences in everyday activities.
Convergence  is a paradigmatic description of the current media landscape, defined as one of “layering, diversification, and interconnectivity of media”. In the past few years, a number of specific conceptual framings have been proposed that deal with convergence and its effects on individuals, organizations, and society. Media studies conceptualizations, such as crossmedia and transmedia, are descriptive in nature and by far the most mature ones; product and service conceptualizations, including those originating in marketing such as multichannel, bridge experiences, and cross-channel, are mostly generative in nature.
Crossmedia is characterized by an emphasis on media-related experiences and on consumption: movies, music, TV. Iacobacci  defines crossmedia as an “environment” where “content is repurposed, diversified and spread across multiple devices to enhance, engage and reach as many users / viewers as possible”. This prevalence accorded to consumption in definitions of crossmedia is supported by Jenkins, who clearly frames it as “a new disruptive pattern of consumption, not of production; the result of the complex interactions of media convergence, participatory culture, and collective intelligence” .
Crossmedia products are expressly designed to be experienced fragmentarily and across a number of different mediums: TV series, movies, videogames, websites, novelizations, and comics. Together these structure a loose narrative ecosystem where each part is deliberately designed to only provide a partial experience, constantly and more or less explicitly referencing and linking to one or more other mediums in the ecosystem.
Crossmedia strives for audience-side completion: references and cues are introduced in the narrative stream to increase perceived complexity and to entice audience members to delve deep into the crossmedia universe itself and so gain a more rounded understanding of it. Overlaps and repetitions are to be avoided .
More recently, Jenkins and others have introduced the term “transmedia”. While offering media studies scholars primarily a more up-to-date term for crossmedia, this is for example Davidson’s position , transmedia ostensibly factors in a novel interest in convergence as a production process, in the mechanics of co-creation, and in the resulting more prominent role accorded to audience agency.
According to Jenkins , transmedia is “a production process”, where “integral elements of a fiction get dispersed systematically across multiple delivery channels for the purpose of creating a unified and coordinated entertainment experience. Ideally, each channel makes its own unique contribution to the unfolding of the story”. This increasingly “unified and coordinated” experience is interpreted by Iacobacci  to be “invasive” and “fully permeating” the audience’s lifestyle.
Increased audience agency offers another way to draw a line between crossmedia and transmedia: “true transmedia storytelling is apt to emerge through structures which encourage co-creation and collaboration”. This is reflected in Jenkins’ introduction of a differentiation between “adaptation”, which “takes the same story from one medium and retells it in another” and is typical of traditional media licensing strategies, or “extension”, that “seeks to add something to the existing story as it moves from one medium to another” and that is a defining characteristic of transmedia . Transmedia is not just about multiple, interlaced, convergent stories, but about creating rich, shared meaning in the space between mediums. For example by adding backstories, prequels to the action, maps of the world, or the points of view of secondary characters.
DeMartino  also explicitly supports this notion of audience participation and engagement. Focusing on the author / consumer relationship within the entertainment industry, DeMartino pragmatically argues that transmedia allows storytellers to “create deeper experiences for their audiences when they unfold a story and its world via multiple venues, and when they invite consumers to participate meaningfully in that world—especially when they do so from the outset of the project”.
The conceptualization of multichannel stems from marketing, where it identifies a way of delivering services that allows users or customers to interact with the providing system through a variable number of different physical, digital and biological channels, thus enabling successful completion of tasks through any of these in a discrete fashion . A financial transaction between an individual customer and a bank could for example be carried out in a number of mutually exclusive ways: through a call center, via an online banking facility or mobile app, or by physically visiting the closest branch office and requesting assistance from staff.
Multichannel is predicated on exclusivity: a transaction cannot be completed while moving across channels. A transfer of money initiated via a phone call cannot be concluded online, or at the local branch, if not through ad hoc interventions. Little to no information is shared between channels, and status is not maintained. Channels have hard boundaries and the system’s awareness of ongoing activities is limited. Rather than offering an integrated, systemic approach, multichannel duplicates and adapts activity flows within parallel but non-communicating processes, and binds them to individual mediums, here called channels. While formalizing the necessity to guarantee that user-system interactions happen in a variety of ways, multichannel ultimately provides a siloed production- or organization-driven approach ill-suited to serve a convergent landscape.
Bridge experiences were originally conceptualized by Grossman in 2006 as a way for user experience designers to respond to the need of “consistently flexible adaptation” brought on by convergence and increased user participation. Agency is Grossman’s main concern, and he maintains that user experience “thrive(s) on remix, mashup, and appropriation” .
Grossman defines a bridge experience as one where “the user experience spans multiple communication channels, document genres or media formats for a specific, tactical purpose” . In bridge experiences, users are “forced or compelled to traverse domains and use a number of different devices, platforms or locations in order to communicate successfully, complete a task, or attain a desired physical, mental, or emotional final state”.
User agency is a core characteristic of bridge experiences: “without movement between domains, a user cannot reach the end goal” . As such, preservation of continuity and elimination of gaps, or seamlessness, is an important concern, together with the “actualization of a specific intent—either of the user or the designer; hopefully both—that demands movement across domains”. Bridge experiences are fundamentally “driven by internal narratives, perceived causal relationships, and interpretive leaps that branch off from one another” and that users “design the overwhelming majority of (them)”.
Also a marketing term , cross-channel originally identified a modality of service delivery where “a single campaign” was driven “with a consistent message (...) coordinated across channels”. It was then introduced to information architecture  and user experience  to describe the changes occurring in the practice in connection with convergence.
Cross-channel user experience design introduces a systemic framing often based on heuristic processes . It refactors the narrative-driven nature of transmedia into complex “user stories” or “user journeys” moving across different touchpoints and mediums, and reappropriates the service-minded, gap-filling framing of bridge experiences to produce an ostensibly more rounded take on contemporary user experience. It is worth noting that cross-channel, while still embracing the idea of seamless, unhindered flow, purportedly supports the notion of visible seams  between touchpoints or channels as navigational and experiential aid, for example to warn users they are moving from a secure to an insecure location.
The diction “omnichannel” again hails from marketing literature , and by and large identifies an alternate term for cross-channel, to the point of being at times presented in trade publications as “cross-channel done well” . Omnichannel has recently gained some traction in user experience practice .
Media studies theories offer mature conceptualizations that can be successfully used to describe and understand convergence in the current media landscape and the way entertainment is produced, consumed, and to some extent co-produced. Crossmedia and transmedia both introduce a narrative-driven, systemic approach to experiencing media that considers audience agency, albeit weighing its importance differently, and the interlacing of story threads across mediums. Nonetheless, their focus is squarely on the entertainment industry. Furthermore, both are speculative or descriptive frameworks whose primary goal is far removed from the factual process of producing entirely new artifacts and from design practice. Arguably, while marketing and user experience framings such as multichannel present their own shortcomings and a generally less solid foundations, they have been long and successfully used in the creation and implementation all kinds of products and services, from banking to healthcare.
The original formulations of bridge- and cross-channel experiences have extended the conceptual solidity of crossmedia to product and service processes and partially addressed those issues that pre-convergence models such as multichannel could not solve. Still, there is a need to account for the larger socio-technical changes occurred in the past ten years and the way they have impacted design practice: the way convergence has been heavily reshaping our everyday activities; the way mobile connectivity has been making these activities part of a pervasive whole; the way this pervasive whole is primarily shaped by anonymous masses of contributors; and the way these contributors freely and individually assemble and disassemble transient semantic environments, cross-channel ecosystems, in order to achieve a desired future state.
To this extent, we propose a more formalized notion of cross-channel ecosystem and introduce a generative approach that includes and expands the seminal descriptions of cross-channel experiences originally introduced to information architecture and user experience in the late 2000s.
A cross-channel ecosystem is the ecosystem resulting from actor-driven choice, use, and coupling of touchpoints, either belonging to the same or to different systems, within the context of the strategic goals and desired future states actors intend to explicitly or implicitly achieve. Cross-channel ecosystems are semantic constructs structured around the idea of “experiences”. They straddle physical and digital space, and include people, devices, locations, and software connected by information flows.
In accordance with systems thinking formulations , the term “actor” is preferred to “user” or “customer” to identify all agents who engage with the ecosystem. Similarly, the terms “media”, “medium”, and “channel” identify different conceptual artifacts. In line with current media studies, “media” here refers to an element of the mass media, global communication, and entertainment industries. “Medium” on the other hand is intended, in accordance with Eco’s critique  of McLuhan’s theories of communication , as a narrower and more specialized concept, that of a “method or system for communication or distribution” of information or, more broadly speaking, content. This interpretation of “medium” allows us to introduce an important distinction with “channel” as a design-specific construct. While a “medium” can be identified with an objective information artifact, such as a movie, a “channel” is a design abstraction crafted for a pragmatic purpose connected to a project’s objective.
Contrary to traditional definitions of a (user) experience as the “the overall experience, in general or specifics, a user, customer, or audience member has with a product, service, or event”  or something that “encompasses all aspects of the end-user's interaction with the company, its services, and its products” , cross-channel experience design acknowledges that actors freely move from product to competing product and from service to ancillary service unconstrained by company boundaries or physical location, constantly creating individual information-based ecosystems that suit their needs and goals. In this sense, an experience is simply any event an actor considers relevant and related to achieving a desired future state.
Such an approach fits naturally to complex, pervasive experiences such as “paying one taxes”, “going to the movies”, or “commuting to the city center” which naturally present fuzzy boundaries and which are the ones mostly influenced by personal mobile computing, social interactions, and convergence.
Organization-wise, this results in a loss of control much more severe than the one Jenkins observed in transmedia storytelling as a consequence of increased audience agency and participation . Design-wise, the change suggests the need for a more strategic and systemic approach that focuses on the overall architecture rather than on the individual points of interaction, as some of these might be inconsequential to the actor’s experience, remain unnoticed, or even be detrimental. As such, a cross-channel ecosystem identifies a space of opportunity for intervention more than a finite artifact that can or should be fully designed or managed.
Resmini & Rosati first posited the expanded role information architecture has to play in structuring a necessary medium-aspecific approach suitable for systemic approaches. In a 2009 paper, they outlined a manifesto for designing “ubiquitous ecologies” whose first point is that “information architectures become ecosystems” . Resmini & Rosati elaborated that today “no artifact can stand as a single isolated entity” and that “every artifact becomes an element in a larger ecosystem” sharing “multiple links or relationships” with other artifacts within the ecosystem. As such, they have “to be designed as part of one single seamless user experience process”.
In 2011, they more precisely specified how the information architecture has to convey an ecosystem-level sense of place, be consistent and offer a degree of resilience, work through reducing the cognitive load and support correlation as a way to expand an actor’s possibilities of action . These are superstructures that precede the design of interactions, and are completely medium-aspecific. Within the span of a cross-channel ecosystem interactions may vary greatly point-to-point, from a screen to a card-reader to a person, so the information being exchanged and its structural clarity is the primary way to successfully support an actor’s activities as she is in pursue of her goal.
We maintain that the most important element of the ecosystem itself is its information architecture, the pervasive structure that links and connects the elements within the ecosystem, since this interconnected set of elements coherently organized in a structure produces characteristic behaviors and, through these, the ecosystem’s own function or purpose .
Accordingly, we consider base constituents of the ecosystem those elements that are relevant to the architecture: actors, the tasks they perform, the touchpoints they interact with, the information channels, and the seams that connect touchpoints and channels. These base elements, described in detail in the next section, are polymorphic in nature, as they can be interpreted differently in accordance with the way actors understand them. Acquisition of actor data, usually through ethnographic research or interviews, is the starting point for any design activity, and the cross-channel ecosystem which is the focus of the design process is usually the compound of a substantial number of individual outlooks. The reason for this is completeness: as individual experiences identify paths along touchpoints through the ecosystem, they may differ substantially. Capturing multiple paths accounts for a larger number of alternatives and a more representative set of touchpoints, providing the designers with a model more closely approximating the actual environment in which the experience takes place.
In the context of cross-channel ecosystem design, the primary elements in the ecosystem are actors, tasks, touchpoints, channels, and seams. An actor is any agent active within the ecosystem. A task is any activity an actor performs towards a future desired state. A touchpoint is an individual point-of-interaction in a channel. A channel is a pervasive, ecosystem-wide information layer. A seam is a threshold between touchpoints and, in certain cases, across channels.
A cross-channel ecosystem is a system of systems: design processes work at different granularity levels, pragmatically defined. A touchpoint could be a smartphone or an app living on the smartphone depending on the current design objectives.
A channel is a pervasive layer carrying information around the ecosystem. Channels are an abstract, high-level construct, and a designer-made artifact: they could reflect the formal sectioning provided by an enterprise architecture model, be the result of user research or any discovery phase, or represent a more informal view of a project’s own context and of the designers’ own biases and interpretation. For example, a channel could be conceived that “holds” the information concerning bus departures in an urban transport ecosystem. Another typical channel might be holding “user-created content”, straddling all of the content created by actors within the ecosystem, regardless of where it appears. As these examples clarify, channels might, and usually do, span across a number of different mediums. In a way, channels can be compared to containers, or pipes, for specific “types” of information and to fuzzy, non-Aristotelian categories.
Individual points of interaction within a channel configure touchpoints. Touchpoints allow actors to receive, produce, co-produce, or modify information. In the urban transport example given above, the individual displays at bus stops, a paper timetable, or a mobile app are touchpoints. Touchpoints are individually medium-specific and can be configured physically (a bus stop), digitally (a mobile app or a software agent), or biologically (a bus driver).
Any two touchpoints might share a seam that allows transitioning from one to the other. Seams can be unidirectional or bi-directional. As information is medium-aspecific and flows through ecosystem-wide channels, seams often connect touchpoints belonging to different channels. In accordance with Resmini & Rosati’s axial model of information architecture , in-channel seams are considered horizontal seams, while cross-channel seams are considered vertical seams.
An imprecise but practical way of describing the interactions between channels, touchpoints and seams is to consider the channels as utility pipes that carry information. Wherever two pipes join, we have a seam. Wherever we have a tap from which you can draw or add information, we have a touchpoint. Touchpoints are where the information contained in the pipes gets to be made available to actors, or where actors modify existing information or create more information to be injected into the flow.
We can provide a brief example of the different elements and the way they relate to each other to form the architecture of the ecosystem. A recent internal project at Jönköping University was started to investigate how to improve the quality of the courses offered to students and was run as a cross-channel design exercise. The research team identified actors first, resorting, because of time and budget constraints, to produce four coarse groups of managers, administrators, educators, and students.
Then a decision was taken to investigate primarily what actually a course “is” to students, them being the primary target and in the context of the project the least represented. Interviews were conducted and a questionnaire sent out. The students, both in interviews and in the questionnaire, were asked to describe what a course was to them, what tasks it involved, and what was their goal: pass the exam, learn about the topic, and so on. In all, sixty questionnaires were completed and returned and ten semi-structured interviews carried out. Results were compiled and analyzed to extract an initial list of tasks and touchpoints, from which the relevance of “unofficial” tools such as Facebook groups or the role of library meet-ups for group study emerged clearly. Tasks and touchpoints were then analyzed on the basis of their role in respect to the co-production, consumption, or remediation of information and an affinity diagram was produced that grouped them in five channels: external sources, including all non-university-sanctioned materials; administrative information, mostly concerning university-level issues; course information, concerned with day-to-day activities and deadlines and mostly managed by teachers; lectures, including all materials introducing the topics taught in the course; peer conversations, including all of the content produced by students across a number of touchpoints, from Facebook to plain paper notebooks.
In the “peer conversations” channel, identified touchpoints included the already mentioned “Facebook group”, but also “YouTube channel”, “class debate”, and “lecture notes”. Some of these touchpoints were further specified during the run of the project. For example, a “FB group – Website” and a “FB group – App” distinction was introduced. Again, these were pragmatic decisions based on project scope, constraints, and goals. Some of the identified touchpoints belong to more than one channel. For example, “lecture notes” belongs to both the “peer conversation” channel and the “lectures” channel, as it is information created by the students (the peers) that is related to the individual lectures. This means that medium-specific touchpoints convey medium-aspecific information across channels and act as a seam between these.
Similarly, by interacting with touchpoints, actors move across channels in the pursuit of their goals, for example by joining together someone’s verbal indications and a list of lectures available on the course learning platform to identify a specific topic they will need to discuss in class. The overall view of the actual ecosystems experienced by the students was then reported back for further integration and design work by means of a number of traditional user experience deliverables such as system maps and user journeys.
In systems theory, self-organization is the result of the activities that form, maintain, protect, and regulate information flows within the system. This set of interwoven relations, actions, and reactions collectively create the system and, in their entirety, constitute its organization . The resulting structure is also the source of system behavior and of system identity . Self-organization is how a system increases its complexity. Increases in complexity produce more resilient ecosystems: that is, more efficient ecosystems capable of better responses in moments of change or duress .
In cross-channel ecosystems, independent actors loosely join touchpoints belonging to different, sometimes competing services, into transient architectures. They do so by walking impermanent but traceable and repeatable sequences and paths that traverse the ecosystem itself, generating emergent structures, creating new information loops, reinforcing existing ones, replacing them when they become obsolete, constantly increasing the ecosystem’s informational content. This way, actors “change or influence the current and future status” of the ecosystem itself, making it more apt at answering specific needs and different seeking strategies , and ultimately more successful at supporting actors reach their desired future state.
In this paper we formalize a conceptualization of cross-channel ecosystems as information-based architectures and propose that such a framing is applied to the design of products and services to better address how actors go through a convergent experience.
We argued that convergence is shaping not only communication strategies in the field of media and entertainment, but also all processes oriented to production, remediation and consumption of products and services, transforming them in co-created experiences where the role of individual actors has greater significance. We examined prominent conceptualizations of convergence, including crossmedia, transmedia, multichannel, bridge experiences, and cross-channel, and argued that designing for a convergent world requires bridging the speculative framings of media studies with the generative methods of user experience at the strategic level of information architecture.
We suggested that such a design approach rests on a four-part key shift that moves: from media-specific experiences to generic everyday experiences. Cross-channel is a generative framing aimed at understanding, mapping, and intervening within ecosystems. Its area of application reaches beyond the media and entertainment industry; from production and participation to anonymous mass co-production. Cross-channel ecosystems are actor-constructed structures whose content is by and large socially and anonymously produced; from tightly scripted organizational control to individually-generated paths. Cross-channel introduces a change in the degree of freedom that actors enjoy in their use of and navigation across channels within the ecosystem; from products and services to experiences. Cross-channel ecosystems are not organization-controlled, nor product- or service-bound. Rather, actors freely move between often competing products and services in order to achieve a desired future state.
We then provided a formal definition of “cross-channel ecosystem” as actor-driven, information-based, semantic constructs connecting individual touchpoints into transient architectures whose primary elements are actors, tasks, channels, touchpoints, and seams. Through their relationships these constitute the pervasive information architecture of a cross-channel ecosystem. We argued that the information architecture is the most important design artifact, as its coherent organization structures ecosystem behavior, function, and purpose, and allows for actor-side sense-making and place-making. We submitted that actor intervention increases the amount of information available within the ecosystem, producing a rise in complexity and a more resilient and successful ecosystem and that the optimization of this process and the adoption of a strategic approach, as opposed to point-to-point intervention on single touchpoints, is the goal of cross-channel design.
We conclude by reporting that projects are under way to more thoroughly proof, test, and improve the framing discussed here.