Models in Motion: Ethnography Moves from Complicatedness to Complex Systems

BUT model assumptions and ethnographic value

The original BUT model that enabled ethnographic praxis in industry to flourish these last 20 years made certain assumptions. Ethnography was interpreted primarily as a way of explaining detailed lives of people. These descriptions often consisted of two parts (Salvador 1999): (1) in situ research to describe the context where a product’s use was or was imagined to be and (2) an emic point of view; the ethnographer could bring back a narratives from the field, ideally people’s own words and actions captured, providing the target user or consumer’s perception (as point of balance to the corporation’s perceptions). Of course there were other foundational parts of the design anthropological approach, like being multi-sited, holistic and generative, though these didn’t have the same weight in terms of value for BUT model. In practice, the result has been to understand contexts and ensure emic voices are heard. These two aspects of context and having the people’s point of view, were crucial in order for the BUT model to succeed at reducing uncertainty around “users”.

The BUT model itself has some assumption that enabled ethnography to thrive with it. First, it assumed that the “market environment” changed at a relatively steady rate. Radical changes or a great deal of volatility in the market were not part of new product development model. Second, the model assumed that the rate was slow enough that classic measures of the market, e.g., segmentation surveys or pricing sensitivity studies, adequately captured a meaningful and relevant state of that market not only at the time of measurement, but at the time the new product would be introduced. Third, as unpredictable as markets for new products are, it assumed that all variables with any likelihood of impacting the new product were known. Therefore, the role of ethnography was to more precisely identify the mean – the product or service that the greatest number of people would choose within the complicated but stable market system.

The assumption in the model is that a company will only take to market ‘a’ product or service, and so finding ‘the’ product or service attractive (or at least good enough/acceptable) to the greatest number of people was a benefit. Now small-scale manufacturing, “makers”, digitization, and so on, allow the tails of the distribution of customer needs to be addressed. So ‘a’ product or service is no longer sufficient. This phenomenon extends from manufacturing (maker movement), to participation (data, virtual) as well as to funding (Kickstarter). More and more companies are finding it necessary to move away from a single product/service strategy to an experimental process (Intuit, Lean Startup).

In fact, one exception to prove the rule is the notion of the “strategic inflection point” (Grove 1999). Grove recognized that markets do indeed change, but that these changes are punctuated moments in time associated with a significant change in the basis or bases of competition. These punctuated moments of disequilibrium can be specifically identified and isolated is demonstrable by these first three assumptions: (1) Of course markets change, but (2) those changes are either slow enough or only occasionally punctuated with significant change, (3) that changes are the exception, rather than the norm.

The main contribution of ethnographic work in this milieu was specifically to qualify social and cultural variables (as opposed to the market variables) that would contribute the acceptance or rejection of a particular product. The ethnographic work would delve deeply into the context and mindset of the people being studied. Ethnography then became a key to unlocking the “U” for putting together the success puzzle in the BUT model.

There is, however, a fourth assumption of this old model – cultural values and practices change slower than the market systems emerging from and riding on them. Even though we recognize cultural values and practices do change, the evidence suggested they change slowly with time – and so slowly as to be effectively and essentially stable and static with respect to new product development. In short, social-cultural change wasn’t assumed to happen in a normal product development cycle. We will re-examine these assumptions in terms of the world today.

New context for ethnography in practice 2.0

Together, these three shifts: the creation of a new class of digital assets, the common availability of commodity parts and skills and the flux of cultural values creates a foundation a very different market system, which itself requires different approaches to comprehend and address the market. We’ve diagrammed this out in Figure 2. The bottom axis is the Velocity of the Market (Vm); on the left, the Velocity of the Organization (Vo); and on the right, the Complexity of the System (Cs). Cs is defined as the total possibility state space. That is, given cultural flux, commodity parts, and digital assets, there are a vastly larger number of combinatorial possibilities of technological means to reach the market and market acceptance; thus, the Cs is a measure of the total potential bases of competition in the market. Vm is defined as the rate of change of these bases of competition, that is, the activity of engagement in the possibility space. Finally, the Vo is the measure of the organizations ability to keep pace with the velocity of the market with respect to the changes in bases of competition. (Of course, “velocity” is a measure of change over time, so time is implicitly accommodated in this figure.)

FIGURE 2. Emergent Market System: Velocity of Market, Velocity of the Organization & Complexity of Market System

Vo has two components, how quickly it can test a specific basis of competition, and how broadly it can probe a complex space of potential and diverse bases of competition, both need reconciliation to find a position of stability or order in the ecosystem. Smaller companies require positions of less order – needing only suppliers and customers (Porter 1998). Larger companies, to maintain larger revenues, require positions of more order to which many companies and bases of competition can orient. Maintaining a position of order is a challenge because it is expensive and risky to invest too broadly or too quickly. To avoid the risk, companies adopt a wait and see strategy – to see what market order or winners emerge. The downside of a waiting strategy is it often becomes too late or too expensive to buy ones way into a new market once the order has mostly emerged with entrenched platform providers in place. Ethnography can help narrow the huge possibility states of potential new orders or market platform winners to a few that are consistent with flux. Further it can guide an organization where to probe or catalyze with early action and investment to optimize the velocity and diversity investment equation.

In Figure 2, we’ve identified the “perfect” or “ideal” position, which is the dashed dark diagonal line, which we assert is the Vo perfectly matched to the Vm and vice versa. We also identified the dashed diagonal line slightly above the dark one, which indicated the “desired” state of any organization to be “ahead” of the market by just enough to gain competitive advantage. Ideally, this is the optimal position for most companies.

Finally, we must consider the four quadrants, which reflect the relationship of Vo, Vm and Cs. The upper left and lower right quadrants represent less of an opportunity space, so we cover them quickly here: The upper left represents an organization whose capacity for addressing the market (Vo) exceeds the rate at which the bases of competition are changing, even though there is high system complexity (Cs). An example of this would be electronic medical records. In the lower right, we have the opposite, high Vm, low Vo, but low system complexity. An example here would be the establishment of a brand new basis of competition that immediately supplants a prior one, and thus, prior organizations. The shift from iron lungs to the polio vaccine is a good example of the clear advantage in a system of little complexity of market solutions.

In the lower left quadrant, we have low Cs, low Vm and low Vo. This quadrant most closely aligns to the “old” BUT model of uncertainty reduction. In this quadrant, the system complexity is manageable, bases of competition are changing slowly enough that classic methods can be used to measure the market activity, detect trends, and plot not only possible, but probably future states with some confidence such that the success of a product release can be somewhat reasonably ascertained, that is, all else being equal. Moreover, organizations operating in this quadrant have the capacity to move with or, if they are really good, be slightly ahead of the market. Ethnography was able to provide the necessary information to enable companies to at least keep pace, and usually be a step ahead of the market. Ethnographic research was the competitive advantage.

However, based on our argument of cultural flux, new digital assets, and new commodity capabilities creating a more complex market system, the upper right quadrant is where all the action is; this represents the “new” model” under which we are operating in a more digital world to develop product, services and create strategies. Our current and old ways of delivering value to organizations in the system need to change. With high Cs, and concomitantly high Vm, we see the bases of competition changing rapidly. Organizations and methods need to adapt appropriately. We will explore this quadrant more fully to understand the system and implications for ethnography.

The high complexity/velocity quadrant

First of all, the system complexity is high – there are more and new digital assets, a commodification of everything and the willingness to apply it. Second, we see a large measure of “disorder” in the system as a result. New products are offered almost daily, from small, medium and large companies, through a variety of market vectors introducing new bases of competition. Each introduction has some probability of bringing a measure of order to the system as other participants orient their efforts to the new introduction. With each new introduction and orientation (or lack of it), organizations at pace with the market (Vo matches Vm) adapt with improvements, changes and/or cancelations.

Figure 3 represents the look and shape of launching new products into a complex system. The system, unlike the BUT or complicated market, presents a very different challenge. The path to success is not a straightforward linear one. It is not a matter of defining the market. All the parts are moving.

The goal of any organization is to reduce the viable/likely options in the possibility space by introducing order to the system. The question is, therefore, how to introduce order into the system, not reduce uncertainty. Uncertainty will remain. This is a crucial distinction; reducing uncertainty suggests understanding the system, per se, prior to launching the product. It suggests a linear, stepwise process. We contend, however, the system is now too complex for a prior comprehension and thus the product launch is itself an experiment about order or arbiter of order. The product launch becomes an indicator of the probability of creating order in the system, even as the system complexity increases with each introduction. Creating order in the system means that at least some of the system elements orient to/around/with the new product/service introduction. Failure to get other parts of the market system to order around the product/service will mean product/service failure. The implication of this is, small companies – or even everyday entrepreneurs – only need to create a relatively small degree of order to “survive” in this system, whereas large companies must create a relatively large degree of order around their product/service to “survive”. It is unsurprising then that Kickstarter now funds about one quarter of start-up activity in Silicon Valley and growing – it enables small introductions to the system that require little to orient toward these products and services. In other words, it is much easier to have small successes, of others orienting toward you, in a complex market than large ones.

FIGURE 3. Product introduction path in a complex market

Eco-system ethnographic studies

Ethnography can play a key role in knowing the systems and enabling agility on the part of the organization in a rapidly changing market that is part of a complex system. There are three immediate implications for this new model as to how ethnography should happened in a corporate environment. First, there is an increased need to have constant on-going ethnography to monitor the social-technological changes occurring. Knowledge, not just methods, is important. Monitoring the social world, in particular becomes important in order to understand what is normative, what is changing, and what might be perceived as an opportunity for stabilization.

We will briefly explore three separate ways some current ethnographic organizations are enabling the ethnographic stalwart of longitudinal or in-depth work to happen. We call these “my people”, “ethno-mining” and “topics” approaches. The examples are meant as illustrative of kinds of approaches.

We look forward to upcoming EPICs for papers that reveal new successful approaches in a complex market environment. The examples here are not endorsements of any of these organizations, the quality of their work or the sustainability of their business practices. We are, however, are suggesting that sustained longitudinal research, rather than frequent movements between topics and people (e.g., domestic help in one project, followed by a project on Chinese youth and cars in another), might not work as well in understanding changes and opportunities for order in a system. We look to these ethnographic research projects as examples of what would support the value of ethnography in the high Vm/Cs realm.

The “My People” model

One approach to ethnography that shows promise is to be deeply engaged, highly specialized, and focus on a particular area of specialty. One example in operation today is China Youthology (http://chinayouthology.com). China Youthology studies one thing – Chinese youth. What is interesting in their approach is the on going monitoring of a particular population segment in a particular geography. China Youthology continuously runs ethnographic explorations with China’s youth. These are classical ethnographic approaches in that they include interacting with the youth in a variety of contexts, on and off line. They are monitoring what youth are doing, how they are acting what they are saying. Further, they are not just distant white coat observers, but also active in supporting and sponsoring youth events in the arts and entertainment. These kinds of activities provide the opportunity for insight not only about youth, but also about companies directly interacting with youth in large business eco systems like technology, arts, entertainment, consumables and media.

The advantages of this approach for our model are a through understanding of the system features relations to that population. China Youthology can understand the macro characteristics of the system and what have been the flux points in the system in the past. The research can differentiate what changes that have really changed underlying social-cultural structures and what are more surface changes. China Youthology has a deep understanding of the people, their values and practices and knows the rationale behind past successes and failures of businesses interacting with the population, and recognizes the emergent nature of youth’s relationship to the market. They are ready to identify and act upon flux areas as they emerge. This is not the only way to do longitudinal ethnographic work with “a people”. Ichikawa (2012) outlined how longitudinal ethnography and collaborative design in a village gave insight not only into the village but Japanese culture. Likewise, Nafus (2013) two year long research with the virtual Quantified Self community, demonstrates that sponsoring community events, conducting collaborative design and being embedded in the community can help ethnographers know not just about the QS community, but also areas of flux in wider American culture. Clearly, there are many ways to do a classic style ethnography with “my people”, following them over time to gain a depth of knowledge about potential flux opportunities.

Big data, behavioral tracking approaches and ethno-mining

Given that the digital is a contributor to the disruption, it is also a source for ethnography to remain vital. We see potential in using digital data to enable ethnography to understand the on-going dynamics of the system. In particular, we’ve seen in the works of people like anderson (2009), boyd (forthcoming), Churchill (2008), Gray (2012) and Patel (2010) are just a few examples of how longitudinal tracking studies reveal larger patterns in the system, and points of disjuncture or flux. EPIC has been a venue that has been open for exploring digital behavioral data not as a threat, but as an important tool in the ethnographer’s tool kit. The approach has similar advantages to the “my people” approach though with a different target population, whether that be on-line communities, users of particular devices, or users of an on-line service. Ethnographically informed big behavioral data enables the ability to notice changes in social and cultural practices – emergent patterns in the system.

Punctuation points – With a high complex and a very dynamic system, it is important to be able to make frequent probes, or tests based on hypothesis about the opportunities for an organization. The “my people” approach or the ethno-mining approach enable a baseline of knowledge of the system. With that kind of knowledge, it is relatively easier to create entry points for these punctuation probes. Nafus (2013) has been conducting these types of probes with the Quantified Self community. These have sometimes occurred in via a regular series of meetings with community members and researchers called “Co-labs.” Further, the QS research project is adding a software app to explore community understanding of areas of interest that might arise, like data sharing, which could impact Intel’s research project around personal data. Another research activity related to this kind of work, perhaps, has been some of the activities of living labs. Although there are many aspects to living labs (cf http://livinglabs.mit.edu) , the key salient feature for us here is assessing new products and services in real life situations, especially the social and market aspects. At this time, we aren’t aware of many other practices in this area, however, these in theory could be similar in structure too much of the work done today with ethnographic practice in industry. Quick studies, done whenever needed, on whatever topic is relevant. The big switch from today’s practices would be the focus. Rather, than a constant exploration of the broad unknown, research would need to target particular hypothesis around social-cultural flux, dynamics of the market or potential ways to create order in the market. These ethnographic explorations are examining both the market and social conditions.

Topics model – Long durations of time is not the only way to understand the breadth and on-going system dynamics. Radka (2011) explained how Claro conducted a consortium research project around ownership. The paper briefly describes the advantages of consortium work in terms of taking on a large question at a reasonable cost in a timely manner. For the purpose of the new design ethnography, a key value in this type of work is the ability to do a series of topic related projects. Following these topics in-depth, doing both ethnographic and business ecosystem work, provides a sense of the system landscape, as well as, what fundamentals are in flux. The model maps well to the kind of product development we envisioned in Figure 3. The ownership consortium project lead to new work on value networks, where another aspect of traditional roles was changing and new forms of value were being created. The value network research then lead to research about the personal data economy. If you recall Figure 3, ownership was entry point into the system. The next series of research projects were pursuing relationships to that entry point to understand where products/services could be to get others to orient around them. The topics approach enables the ability to pursue topical relationships in a system to understand which parts of the system are malleable for stabilization or disruption.

Ethno-analytics

In the old BUT model, ethnography was primarily focused on the detailed understanding of people and their contexts. With a new complex market model with a complex Cs and increased velocity Vm in which ethnography needs to add value, there are new opportunities for us. A key aspect is contributing to understanding the market place, not just people and their contexts. We have conducted two different experiments utilizing ethnography to provide insight for our organization to take action in the market of the emerging data economy. We are going to use examples from our recent exploration into data economy. The space of interest around the data economy emerged from flux work around changes ownership, everyday entrepreneurships, new forms of participation, changes brought about mobility and the rise of new sensing technologies. We developed a vision of an emergent data economy, in which individual’s data circulates, combines with other data (public and private), and returns economic or participatory value. The premise was on an individual’s ability to access and use their personal data.

Eco-system patterns – Ethnography provided a broad understanding of the space for a data economy. Still, we needed market data to determine if there was evidence of disruption of market opportunities. We needed evidence to suggest that “data” is becoming the fundamental driver of the digital ecosystem (vs. traditional drivers of device performance, power, form factor, etc.). Partnering with QUID (www.quid.com), we used the criteria for data economy model from ethnographic data, to run a big data analysis against publically available information about companies. What we were able to create was a history of the rise of a data economy from 2000 until the time of the study in 2011. We were able to map the dynamics and relationships in the business ecosystem that could form a data economy. The analysis revealed that by 2011, private investment in data companies exceeded investment in traditional computing companies (See Figure 4). The companies in the circle were entirely new since 2000. Furthermore, on its current trajectory, other “computing companies”, though perhaps remaining profitable, may lose “centrality/ influence” in the broad high tech industry unless they secure strategic positions in the emerging data economy. Figure 4 represents a historical snapshot of the ecosystem in 2011. The digital ecosystem, as traditional computing companies (semiconductors, networking, storage, etc. represented by darker dots) are repositioned in the global digital ecosystem network as data-centric companies (lighter dots, named ‘vibrant data’ companies) are funded. The ecosystem mapping provided a big data and visual analysis that showed an opportunity space, as well as, some key clusters that could form the basis of gaining footholds of control or strategic influence in an emerging data economy. The main point here is rather than traditional representations and descriptions being used to create something like “personas”, the ethnographic content is used to explore the market dynamics.

Figure 4. Ethno eco system analysis

Eco-system simplicity from complexity – Although we identified the data economy as a potential market because of social-cultural values in flux, we needed to understand the potential of the space. How can a disruption present opportunities for a company (e.g., Intel) to provide the technologies and platforms that empower individuals to benefit from the secure circulation of data, creating and exchanging value to the benefit of themselves and their communities. To narrow the scope of the huge opportunity space, we engaged in an expert sourcing (like crowd sourcing but with a very select crowd of experts) of the complex network of factors that could lead to a data economy (www.WeTheData.org). The analysis itself was not done in the typical ethnographic style, but through an open crowd-sourcing process. Ethnography, however, was crucial to the process. It was through ethnography that we could identify the 90 variables for analysis. From the 90 variables, network analysis showed a cluster of 22 factors that might be considered “leverage points” or potential points of order in a system – these are the catalysts in the market on which we could focus our attention, either by developing technologies for the marketplace, enabling an ecosystem of developers, or other means. Collectively these factors pointed to four broad “challenge areas” or “design criteria” that would have the highest likelihood of catalyzing a new data economy: Digital Trust, Data Literacy, Platform Openness, and Digital Infrastructure (see Figure 5). These serve as a foundation for potential actions internally as well as externally. Further, by making the research and analysis part of an open innovation system, Intel has been able to catalyze and direct fellow travelers and entrepreneurs into the market in a way that furthers Intel’s path to successful entry. Ethnography can directly contribute to creating market directions for organizations, just like we have in the past toward product development. The techniques are new and different, but ethnography’s ability to contribute value in a market environment like we described of fast Vm and high Cs.

FIGURE 5. Key factors that emerged from network analysis: Platform Openness, Digital Trust, Data Literacy and Digital Infrastructure

Ethnographic teams

As should be evident from the changes in how we work and where we should focus, there is equally a shift in who we should be working with on our teams. Designers continue to play a key role in making results actionable. However, there are other key new partners in the ethnographic team. Big data analysts play a key role in helping to capture the complexity of both the market environment and the patterns of cultural flux. In our example above, the scouring of the web to help understand the transformation and emergence of a data economy ecosystem relied on big data analytics. The partnership between big data and ethnography was essential in each being able to properly deliver results. In terms of ethnography, papers like anderson et al. (2009) and Patel (2010) have shown that big data can be a partner in the ethnographic enterprise, especially in a longitudinal approach and discovering patterns of cultural flux. Closely related to the analyst, is the data visualization expert. The visualization of the data has a key role both among researchers, as well as, representing the data out to corporations. The dramatic visualization of the unfolding of the ten year story of the data economy market was a compelling. It created a powerful visual story. This is more than “information architect” – it’s creating various forms of meaning by understanding data otherwise not understood – visualization creates the understanding as well as describes the data. Finally, bringing business analysts, e.g., economists, directly into the process become critical in enabling an organization to take understand and take action in the market. It is a good sign that many ethnographic firms that have presented at EPIC already have business analysts on staff. In the examples we have presented here, a business analyst has been integral to all of the research. In particular, business analysts are key in enabling a space for an organization or company to create order, as well as, creating and experimenting around catalysts.