XXXI.2 March - April 2024
Page: 28
Digital Citation

Customer Experience in Telecom Service Design: Time to Up the ANT

Ruth Neubauer

back to top 

This past year, I was surprised to find myself having a really bad customer experience with my mobile Internet provider. As a customer, I always try to draw lessons from the experiences I have, as they inspire and inform my own design work. On this occasion, I found such an intriguing mix of things that could—and did—go wrong, that it inspired me to write this article. My intent? To make the best of this miserable situation and pass on these lessons to current and future customer experience (CX) designers.

back to top  Insights

Actor-network theory (ANT) helps designers to understand the full customer experience.
Internal business processes are active contributors to customer experience and need to be analyzed in the design process.
Visualizing human and nonhuman actors (actants) in a service system reveals potential black boxes and issues that may then be addressed.

It is worth noting that this experience did not take place in the U.K., where I used to live, but rather in another European country where I now reside, where regulations are perhaps less strictly followed. And I do not want to single out this one telecom company, which I deliberately do not name, as I believe that all telecom companies face similar challenges. Nevertheless, I believe it is a useful exercise to record the lessons learned from this experience. I am also using this as an opportunity to offer a demonstration of the power of using actor-network theory (ANT)—a tool from sociology and philosophy, founded by a group of scholars around the late Bruno Latour [1]—to highlight areas for design interventions. ANT follows a school of thought where all social matters can be explained through pointing out the direct relationships between actors—people and things—in a network. These actors may be human or nonhuman; Latour calls them actants. Human experience can thus be explained as an effect of the network of actants. Experiences surface through looking at the actants in a system. An ANT visualization helps us trace flows, fractures, and where things go wrong—or at least not as planned—in the interactions between an organization and its customers.


I thus traced the design of the service I experienced with designer eyes. Puzzle piece by puzzle piece, I completed the picture of how I experienced the flow of actions and interactions. Each part probably made sense from the perspective of the team who created it. I understood what was happening as a designer, but I did not understand it as a customer. As a designer, I could see how the system was intended to provide good customer experience, with a considerate and mostly empathic help desk staff. As a designer, however, I could also see how they failed in their aim. There was a conflict in what I was trying to accomplish as a customer and what the design and engineering team who created the experience were trying to accomplish. Having worked in companies like these, I know it is difficult to coordinate one team's work with another team's. But such coordination is required if the ultimate aim is a good product or service.

I thus traced the design of the service I experienced with designer eyes. Puzzle piece by puzzle piece, I completed the picture of how I experienced the flow of actions and interactions.

One of the key elements in design is the visualization of the components of an interaction flow or set of flows. Visualizations hold important clues for the implementation of an idea. Specifically in service design, it is good practice to use something like a customer journey map or a service blueprint [2]. In my experience, ANT can support this process, as it draws attention not only to what users and service staff do but also to the effects of the interactions of different actants—the human and nonhuman actors, as I point out above. ANT shows how networks of people and things are "programmed."

Figure 1 shows an example of an ANT visualization. This visualization shows how the "program" and the "antiprogram" of a hotel service, as the various actants, relate and thus influence or shape the interaction flows. Again, it is important to note that an actant can be human, like a hotel manager or a customer, but also an object, or a verbal or written notice. In ANT, interaction flows emerge from the relationships within a network of action [4].

ins02.gif Figure 1. ANT diagram by Madeleine Akrich and Bruno Latour [3], redrawn by the author.

This particular story goes like this: The hotel rents out rooms by giving customers keys. The company generates profit through doing this over and over. This requires the customers to return their room keys. This is the company's "program." The forgetful customers' "antiprogram" finds that it is too much work to remember to return the keys, so customers wander off with their keys. When the manager reminds them (2), it has some, but little, effect. When the manager places written notices around the hotel (3), it has a bit more effect. But the company's program really gains traction when a metal weight is added to the key (4) so that customers find it more cumbersome to wander around with a heavy weight in their pocket than to remember to return the key. This example shows the agency of nonhuman actors in influencing the unfolding of the interaction sequence. The flow of action and experience is dictated by the embodiment of the heavy object that makes acting any different from returning the key really inconvenient.

As this example illustrates, an ANT visualization includes how groups of people are programmed through actants in the network and how programs may conflict and work against each other. Of course, there may be any number of groups (and their) programs meeting in an encounter. It is a matter of visualizing this encounter so it can be analyzed and changed. Here I should note that as a one-team researcher and designer in this case, I can only rely on my personal observation in this analysis—technical or operational details may be wrong. However, the reporting of my experience remains on point, even if I have somewhat simplified the details of the encounter. What I present here is what I pictured during the experience. I also demonstrate step by step the creation of an ANT visualization, each time adding more programs and actants to capture the full experience.

Figure 2 illustrates my program. I was working remotely via my mobile Internet.

ins03.gif Figure 2. Actants participating in my practice of working remotely.

Simple. But this visualization could not account for the issues I encountered. First, we need to picture the whole experience for working remotely. I need to add my mobile provider as another program (Figure 3).

ins04.gif Figure 3. Adding some more actants for a more comprehensive visualization.

This picture is more realistic. Not least because who could build their own Internet infrastructure? Or who would build the infrastructure for free for others to use? Mobile services have become a commodity. This was the picture I have always had of my mobile Internet service. But suddenly, my Internet connection was suffering. On some days the service was so bad and the data traffic so low that it was virtually unusable. I was puzzled, as my technical understanding—my mental model—was that there either was or was not a connection. This virtually-dead-but-not-quite state was unusual to me. I have certainly experienced complete loss of connection in power cuts or storms. I also know how it feels when lots of people share the same network—everything goes very slowly. But this was different. I shrugged off my provider's offer to upgrade my contract with more bandwidth, as it didn't feel like a bandwidth problem. Neither did I want to commit to a new contract if I didn't know what the problem was. But I was in the difficult situation that I needed the Internet for work, and that I was so remote in the countryside that my Internet provider choices were very limited.

I decided to investigate and tried to speak to the technical team at my provider. There, I learned that different contracts had different types of access to the network. For example, if someone paid for more bandwidth, they would be prioritized. Or if someone was with a subcontracting (often cheaper) provider on the same network, they would be deprioritized.

This prioritization business was new to me, and it also appeared to be mysterious like a black box. It absolutely made sense to me that companies have business priorities and need algorithms to organize their processes in some way. The diagram shown in Figure 4 represents these business processes. From the outside, however, they remain unexplained.

ins05.gif Figure 4. Visualizing more programs and actants, uncovering business processes.

The technical team also said that my modem was an older model, so I thought perhaps it was broken, or the modem hardware was outdated by new software in the network system. I decided to try a new modem, and accepted the company's offer of a replacement for the one that it had supplied initially. I had figured out that devices supplied by a company need to be amortized through monthly fees. My 24-month introductory period had just passed. I was now on a "rolling contract," still paying the same amount monthly. Perhaps naively, I thought that the replacement modem was included in my new monthly rolling contract. Having proved that I was a reliable customer, my expectation was to get this level of service. Instead I discovered that I was locked into another 24-month contract. I was not sure how it had happened, and I did not remember agreeing to it. I felt I had been tricked into a contract. "Dark patterns" were a thing back when I was designing user interfaces, a thing that was frowned upon in the designer community [5]. Now I had been caught by dark patterns. I am sure no designer had intentionally designed this service trap. Nevertheless, it was there. Perhaps it should have been the job of the designers to prevent it.

There are business and design tensions. I remember as a designer I used to have discussions with the sales team about prioritizing the customer experience or the revenue target for each month. Often, management was on the side of the sales team rather than that of the design leads, focused as they were on landing quarterly budgets. I often had to stand by and watch as they would start upsell campaigns and other measures to accelerate conversion rates, while customer service issues were deprioritized. This did not fit with customer lifetime value and customer first, which are key priorities in our ethics and practice within design.

I realized that the different teams within the company overruled one another, and that my experience changed depending on who got involved.

I created yet another version of my ANT diagram (Figure 5). Perhaps you notice that these are increasingly bigger visualizations. In each iteration, I add a program for each department that gets involved in my quest to work effectively remotely. In this journey, each program adds more actants that participate in my story. Actants that facilitate my journey, but that also complicate my journey.

ins06.gif Figure 5. Adding another program, revealing some actants that complicate my journey.

I insisted on involving the legal team at the company, and I managed to speak with someone who assured me that it was within my rights to get out of the contract. As I had not formally agreed to the new contract, it would not be binding. I was so glad; the mistake had been realized. This led to the diagram in Figure 6: The legal team had overridden the contract that was put in place by a sales executive.

ins07.gif Figure 6. Internal programs of action override each other.

I waited for a formal notification that my binding contract had reverted to the previous state. Nothing happened, though, and I contacted the company again. I was told that I needed to return my modem for the process to proceed. I immediately went to the post office to send the modem back. I was sure this was the end of the story, but a few weeks later the modem was returned to me, with a note from the logistics team that the 14-day period that would allow me to withdraw from my contract had long passed. I was certain this was a misunderstanding, that the logistics team had not read the latest notes from the legal team. The legal team had confirmed that I was within my rights to be released from the contract, regardless of the time period that had passed.

I realized that the different teams within the company overruled one another, and that my experience changed depending on who got involved. This last diagram (Figure 7) shows how the 14-day rule locked me into the contract again. I had received two conflicting messages from two different departments, but it was the 14-day rule, which the logistics team had invoked, that dominated all following conversations with the company. I continued to be locked into the contract for a few more months. Every time I called the customer service line, the agent was confronted with the two conflicting notes in my customer records, and they were not able to act.

ins08.gif Figure 7. Conflicting programs of action become visible.

This is a service analysis, and a multisystems and multiorganization journey—all experienced as one company from my point of view. I would like to illustrate with this increasingly bigger diagram, that as designers we can try to be more complete in visualizing customer journeys, by thinking about the different teams that are involved in providing a service, their incentives and motivations, and how together they affect customers who are moved from one team or division to another. Even the interaction between different teams in the organization affects the customer. Organizations are networks of actants who interact. As illustrated through my story with the mobile Internet service, parts of this interacting network may have conflicting messages. These competing programs of action and their outcomes have effects on the customer. The ANT visualization might help to better understand the effects of the company interaction and balance the outcomes. The optimal customer experience might be an ever-elusive aim, but through paying attention to balancing the outcomes as designers, we might come close.

In my case, I was a technology-savvy customer who nevertheless ended up in a miserable situation where I negotiated for months with the company to get out of the contract. Almost as a parting gift, in the final letter I received an explanation for my Internet issues. This could be filed as a joke letter. The letter let me know that out of goodwill the company would release me from the contract. The letter invoked the concept of "cellular respiration"—a biological concept to explain metabolism and the continuation of life. The company let me know there was not really much it could do amid the wall constructions, surroundings, and weather that all influence the natural fluctuation in connectivity. Despite all this eloquently shared information, I never learned why my Internet connection sometimes stopped almost completely. Even the new modem that had triggered the new contract did not improve the situation. As randomly as it occurred, it stopped occurring. My analysis: This was not something natural but something in the programs of interaction between me and the company, something that was hidden from me as a customer, and probably also hidden from the designers within the company.

I turned to ANT. Time to up the ANT and begin visualizing the processes that produce these natural, explained away, phenomena of poor service design.

This is my call to action to current CX designers: Try ANT as a mapping tool for the interactions within your organization. You might discover some black boxes that need unpacking. You may discover organizational boundaries and incentive misalignments that will lead to very poor customer service delivery. You may work across organizational boundaries with teams, and it is important to know that these teams within the organization are your colleagues in the delivery of excellent CX. It is your job to find out what effects a company and its processes has on the customer experience. And who knows, the ANT visualization might render visible some of your customers' issues that are currently not being raised, and that could not be explained without a clear mapping of the network of interaction across the organization, and its potentially competing programs of action.

back to top  Acknowledgments

I would like to thank Elizabeth Churchill for her conceptual input, which really helped me bring this article to life, especially her thoughts on organizational relationships and how they influence the actor network and thus the customer experience.

back to top  References

1. Latour, B. Reassembling the Social: An Introduction to Actor-Network-Theory. Oxford Univ. Press, Oxford, U.K., 2005.

2. Kalbach, J. Mapping Experiences, O'Reilly, Sebastopol, CA, 2016.

3. Akrich, M. and Latour, B. A summary of a convenient vocabulary for the semiotics of human and nonhuman assemblies. In Shaping Technology/Building Society: Studies in Sociotechnical Change. W. Bijker and J. Law, eds. The MIT Press, Cambridge, MA, 1992.

4. Wright, P. Reconsidering the H, the C, and the I: Some thoughts on reading Suchman's Human-Machine Reconfigurations. Interactions 18 (2011), 28–31.

5. Brignull, H. Darkpatterns.org: Naming and shaming sites that use black hat, anti-usability design patterns. 90 Percent of Everything. 2010; https://90percentofeverything.com/2010/08/16/darkpatterns-org-naming-and-shaming-sites-that-use-black-hat-anti-usability-design-patterns/

back to top  Author

Ruth Neubauer is a university lecturer for design at New Design University. Prior to that, she worked as a designer for public sector technology services and for one of the "Big Six" energy companies in the U.K., where she was responsible for UX. [email protected]

back to top 

Copyright held by author. Publication rights licensed to ACM.

The Digital Library is published by the Association for Computing Machinery. Copyright © 2024 ACM, Inc.

Post Comment

No Comments Found