Visualizing tensions between perspectives is a starting point for uncovering voices and creating meaningful change.
By Julia Beard, Quin DeVries, Naman Mandhan, & Stacie Sheldon
In physics, tension is described as the force that occurs when a rope or similar item pulls on an object. There are two forces at opposite sides of the rope creating the tension as depicted in this photo. Source.
What and When
Let’s imagine that we are interacting with research data and we discover two very conflicting points of view about the same topic such as, “I drive an EV to help combat climate change” vs “EVs are not practical and I never want to have one.” When confronted with data that reveals these moments of tension, we have a choice as practitioners to either ignore that tension (not the right option) or further explore the data to identify how we might loosen this tension. This is when the In/Tension Modeling method comes in.
In/Tension Modeling can be used throughout the design processes, but is often utilized during the synthesis phase to identify, track, and visualize data that is in conflict with one another. This method can be used for creating unique personas, communicating spectrums of user voices to stakeholders, evaluating how to shift those voices with strategy and design, and so much more. By the end of this article you will be able to clearly understand and communicate a spectrum of user voices for your stakeholders.
Why In/Tension Modeling
As UX practitioners, we are called upon to communicate and often quantify the spectrum of human motivations and behaviors as an outcome of research, but how we do that is up to us and there aren’t a lot of tools or resources for this. In/Tension Modeling can capture this spectrum and go a step beyond that to inform business decisions. But that is just one benefit of this tool. Other benefits include:
By providing a structured approach to leveraging research to create qualitative personas, In/Tension Modeling can lend more legitimacy to those personasUsed as a vessel for storytelling, In/Tension modeling can be used to communicate the value of qualitative research to multiple stakeholders and help them find common ground when faced with conflicting viewpoints.It can provide product teams with a starting point on how to approach their product strategy, and use it as a tool to build out artifacts such as product roadmaps, that are rooted in robust qualitative research.It can amplify the softer middle voices between two conflicting points of view and then lead to innovation for those voices
Organizations often wish for their personas to be data-driven, and even sometimes struggle to find the value in qualitative data for their business purposes. In/Tension Modeling can help persona voices emerge from the data itself. Going back to our earlier example with EVs, perhaps you have a group of users who tell you that they would never consider purchasing an electric vehicle because it doesn’t fit their lifestyle or needs, or they simply dislike the change. On the other side of the coin, you have people who tell you that we are destroying the planet, and it is our responsibility to invest in electric vehicles. And then you have other users somewhere in the middle of that spectrum whose voices are perhaps not quite so compelling to memory and are more easily lost in the shuffle; however, they represent a large quantity of people and qualitative data points.
Imagine a two-sided weighted scale. In/Tension Modeling lets you place your most polarized voices at each end of a scale (there are two personas right there) and then lets you see what is probably a surprising amount of more middle ground (another persona?). If you are synthesizing in person you could move your sticky notes onto lines you draw to represent the scales, or if you are synthesizing virtually you could do the same in Miro or Mural or similar tools. Now you can look at their behaviors and pain points and think about how to move those voices toward consensus by putting them at either end of the scale or somewhere in the middle.
An In/Tension scale allows us to place polarizing voices at opposite ends of a scale, with softer voices in the middle that might help connect the two conflicting viewpoints to each other. Source: Naman Mandhan
It is in our nature as humans to be drawn to the most extreme points of view or loudest voices. This can lead to biased research findings or reporting because those extreme points of view are so memorable. But this tool helps us to see there is plenty of opportunity, opinion, and sometimes consensus in the middle.
While it may seem easier to focus on the loudest voices at each end of an In/Tension scale, the spectrum of voices that exist between these loud voices provide an opportunity space on how to bridge the gap between polarizing views. Source: Naman Mandhan
Who are we designing for? Let’s look at our personas
These days, personas take place in many different forms and under many different names. Archetypes, proto-personas, qualitative profiles, and just “personas” are used to identify who we are designing for and to what degree. However, being complex beings, we also have opinions, thoughts, beliefs, and needs that can not only oppose other’s, but also sometimes our own. This is where creating personas can get tricky. If two users have the same goal or are doing the same tasks, but have different needs, how might we better identify and understand the needs of each user, yet also account for the tradeoffs that come with it?
“How might we better identify and understand the needs of one user, yet also account for the tradeoffs that come with it?”
Inspired by the urban planning work of Richard Saul Wurman in the 1970s, The Understanding Group uses In/Tension modeling to document and visualize stakeholder priorities on a continuum to help inform architectural design and strategy alignment. Then they use voting sessions to identify where strategy and business tensions (or consensus) arise in the group and show the stakeholders’ anonymous votes as a visual to see where their opinions/priorities/thoughts lay among the group.
Conducting workshops and voting sessions are one way UX experts collect feedback and gather alignment, but they can be time and resource intensive. A more common method of collecting information is conducting stakeholder interviews. While conducting our own interviews, we saw an opportunity to evolve TUG’s concept of In/Tension modeling beyond creating real time alignment among stakeholders, and to use the technique to discover and visualize how data points across interviewees and across the research session can vary. During interviews, we noticed that there are always trade-offs and that these trade-offs are often not thought about by users. For example, in a modern digital experience users often want customization, and yet they also want privacy. Common in persona creation, these trade-offs often become the sacrifice of one user and a win for another, creating tension within a persona.
Implementing In/Tension modeling for a law enforcement agency helped us to uncover a ripple effect — when you solve one problem, other problems are naturally created. Our client experienced tensions when they implemented a take-home car policy in their agency. The intent of the policy was to give employees more flexibility in their schedule by allowing work cars to be taken home at the end of their shift instead of dropping them off at the station. During interviews a ripple effect emerged; the younger employees loved the perk of getting home sooner, while more senior employees thought it was a threat to culture and comradery. Using In/Tension modeling during our analysis allowed us to map out these layers of a decision and ask ourselves how we might balance out the In/Tension scales without hindering their progress, while also maintaining their spirit or sense of camaraderie.
These trade-offs often become the sacrifice of one user and a win for another, creating tension within a persona.Above is a visual of the take home car example mentioned earlier. Research uncovered two opposing sentiments, “I want to take my car home from work every day” and “I want to go into the station to pick up my car.” The arrows show when one decision is made, there is a trade-off (or at least a compromise) that must be made. If employees can take home a car every day, they can get home sooner; however, the trade-off or compromise must be made to not have as much time spent with your squad with the possibility of not even seeing your squad each day. Source: Naman Mandhan
Asking ourselves ‘how might we’, paired with In/Tension modeling, gave us the following insights:
Insight #1: Through research, we were able to hear multiple points of view on how policy has affected employees. For some, it enhanced work-life balance, for others that same policy hindered camaraderie. Through In/Tension modeling, we were able to track these opposing insights, helping us to uncover that take-home cars are not the problem that needs to be solved. Rather, the real problem that needed to be solved is, how might we maintain camaraderie while still implementing the take-home car policy.
Insight #2: In/Tension scales provide us with a holistic perception of progress. As solutions to our ‘how might we’ questions were tested, we could start to see the scales even out. Helping the organization and its stakeholders see take-home cars was just a ripple effect from the real problem.
Source: Naman Mandhan
Going a step further
In our law enforcement project, we found the In/Tension modeling to be an effective tool for communicating divergent points of view and addressing the ripple effect. In a subsequent project, we were not only able to scale this tool fourfold, but also leverage it to increase stakeholder engagement with the work. These scales provided a visual and approachable tool that brought along product and UX customers in synthesis of the findings and building of qualitative personas from interview data over 120 customers in 6 different countries.
The team started off by conducting the interviews with the goal of understanding current attitudes and behaviors towards vehicle technologies and features. As expected, we found several differences in how customers used and perceived the technologies in their vehicles, owing to variations in cultures, socio-political factors, and infrastructures. However, we also gained insight into several commonalities that existed between customers from different countries — these hinted at the core human beliefs that transcended the boundaries of countries.
As we gathered more voices, we started unearthing themes that the voices fell into. These included beliefs around the role of a vehicle in a customer’s life, attitudes towards privacy, and how customers felt about the future of autonomy in vehicles. The deeper we dug, the more we uncovered conflicts that existed within these themes — while some customers expressed a desire for more autonomy in vehicles, others didn’t want anything to do with it. While some didn’t pay attention to privacy, others wanted full control of it. Additionally, we uncovered a softer, but just as important middle voice. These were the voices that could see themselves in a future with autonomous vehicles, as long as certain practical considerations were met, or ones who understood what they were gaining when they gave up their privacy.
In/Tension modeling added structure to these conflicting viewpoints and visualized our understanding of the spectrum of attitudes and behaviors that customers had across different categories.
Source: Naman Mandhan
And this unlocked something truly powerful.
Unlock #1: It built a bridge between the qualitative nature of research synthesis and persona building, by providing an intermediary step that illuminated the conflicting viewpoints that can exist between personas. Now, instead of showing two distinct personas with conflicting viewpoints, we were able to show our customer how and why these personas came to be.
Unlock #2: It provided a path for Product Managers to see what solutions could help move the needle from one end of the scale to the other. For a Product Manager who was responsible for bringing more autonomous vehicle features to the road, it showed what practical considerations would help accelerate adoption. For the Product Manager who was responsible for the governance of privacy in vehicles, it illuminated what factors would make someone more accepting of technologies by addressing their privacy concerns.
Unlock #3: It allowed personas to become actionable and dynamic. Each In/Tension scale becomes a spectrum to identify each persona’s attitude toward a particular product. This keeps personas dynamic by tracking the possible evolution of each persona and identifying opportunities to support their transition of behaviors/mindset.
The solution to one problem is simply the creation of another. Because humans are inherently complex, there will always be something on the other side of the scale. As UX professionals, our goal is not to be right, but rather; less wrong. The best way to be less wrong is to continue identifying the ripple effects of a solution. That’s why using In/Tension Modeling is so powerful, especially in the analysis phase. For both automotive and public sector clients, this methodology helped us to visualize and communicate these In/Tension moments, ensuring that all perspectives were heard and represented in a variety of personas.
These visuals not only aligned us as researchers to identify different personas but also helped us recognize that conflicts exist within personas, holding true to the human condition.
However, that’s not where this method’s powers stop. Gaining insight into these differing perspectives, through personas, allowed stakeholders the opportunity to identify their larger goal and uncover the gradual steps needed to guide each persona to that goal. Paired with continued research this methodology also has the power to detect when some personas may have taken a few steps away from the larger goal, allowing stakeholders to identify the level of success of a product and assess the appropriate steps forward.
Harness the power of In/Tenson Modeling in your UX Work
In your reports
Let’s imagine that you are doing some stakeholder interviews within an organization about their Intranet. It is not just a fun or shopping website, it is an essential core tool for their organization. You find that stakeholders are willing to open up during the interviews and are straightforward about not liking change and not wanting the website to change. And on the other hand, you have people openly disparaging it and wishing it to change. Both of these voices are interesting and memorable. And when sharing research readouts it can be tempting to only note these strong voices. But as you can see here, when we really looked at the results and loaded them onto the scale, plenty of middle ground emerged. And this itself is an interesting story to share with your clients. It can also help identify opportunities for considering the impact to those who are resisting change and how the experience of a redesign could accommodate their needs.
Source: Naman Mandhan
In workshops
When working with participants in a collaborative persona building workshop, a combination of In/Tension modeling and creative storytelling can create a bridge between qualitative data grouped into themes, and the distinct characteristics that form a persona. Instead of looking at a theme as a simple grouping of qualitative data, In/Tension modeling leverages the power of storytelling to transform these groupings into points of conflict expressed from the perspective of a persona. When expressed in this form and added to an In/Tension scale, grouping of qualitative data becomes something that all humans can relate to — conflict — which in turns helps participants better grasp how different personas can be informed through unique combinations of distinct, conflicting characteristics. You can see from this image below that the models visualize different stories and levels of conflict.
Above is an example of how In/Tension modeling might appear during synthesis. There are several scales representing qualitative data that are in tension with one another. These scales are positioned beneath a dark blue box, indicating that they have been sorted into similar groups. For instance, there might be a blue box labeled ‘Technology Attitudes’ or ‘Learning Styles.’ Additionally, some of the boxes on the scales are of different colors, used to highlight any connections in data from other scales and to identify compelling quotes that could be helpful in the future. Source: Naman Mandhan
In your strategy
Once qualitative personas have been created and socialized, the next step is to use them. The following In/Tension method is meant to build empathy with the various personas affected by a particular product, service, organizational change, etc.
The idea here is that you will use a spectrum with two opposing variables on each side; the middle of this spectrum is neutral ground. Product teams will then plot where each persona lands on the spectrum based on the information provided. This method allows product teams a visual representation of how what they create could affect each persona. This is always a great way to check for bias during the design process; is the decision you’re making for you or for the product and customer’s best interest.
Source: Naman Mandhan
Takeaway: A Practical Step-by-Step Guide for In/Tension Modeling
Conduct qualitative research using your favorite (or most applicable) qualitative research method (although we have a hunch this will work for quantitative data too, but maybe that’s an article for a different time).Study your qualitative data and code it. When creating personas, identifying goals, frustrations and jobs to be done is a good starting point, but feel free to tailor this with additional codes that match your research objectives.Conduct affinity mapping to identify themes and use those themes to create persona voices.Identify conflicting persona voices and put them on two ends of a spectrum. For example, “I see a future where a vehicle drives me everywhere” and “I never want to give up the ability to control my vehicle” are two voices in conflict with each other.Continue identifying conflicting voices. You may discover a middle voice that fits between the two extremes that you identified. For example, in a world of autonomous vehicles, a middle voice might be “I want the vehicle to drive me when I want it to, while still being able to take back control whenever I want.”Congratulations — you have now conducted In/Tension Modeling!Grab a drink. This was hard work!
Tension modeling as a UX research and strategy tool was originally published in UX Collective on Medium, where people are continuing the conversation by highlighting and responding to this story.
