Solving Mobile Challenges with Psychology-driven IA

Posted by
“In privacy-related issues, our mission as IAs is to make the user feel safe. We are dealing with the human mind and emotions, and it is therefore imperative that we understand what makes people feel safe, and what is important to them as users.”As the field of information architecture matures, we are beginning to understand the new challenges it raises for wireless media. This article suggests that some of these challenges can be best addressed through an approach called “psychology-driven information architecture” (PDIA), which bases design decisions and solutions on the psychological profile of the end user.

As an example of how psychology-driven IA can be applied to wireless applications, the article focuses on a service called friendZone, a commercial, location-based community and communication service available across Europe and Asia. Specifically, I will demonstrate how a psychology-driven design approach can address one of the predominant challenges of the wireless user experience: ensuring privacy. Along the way, I will illustrate how psychology-driven IA can be applied to the design of any interactive medium.

The challenges of wireless IA
When dealing with wireless IA, there are three main limitations that first come to mind: limited screens, limited screen display space for links and data, and limited bandwidth. As technology advances, it will grant us higher bandwidth, more screens, and sufficient screen space for all the relevant content we want to present (as we can already see in handsets such as the XDA Pocket PC). However, these are technological solutions to technology limitations. Even with greater numbers of higher-quality screens and more bandwidth, there are other significant conceptual issues that will not be resolved simply through technological advances. Privacy is one such issue.

Privacy problems appear in almost any interactive, networked medium like the internet or interactive television. But in the wireless medium this problem is even more acute, for several reasons:

  • It’s always on: The mobile phone is always with the user and always connected (to both voice and data networks).
  • It’s personal: The wireless medium enables each user to be uniquely identified by his handset, unlike the internet, for example, which on a home computer can be accessed by several users.
  • It’s location-aware: The wireless medium is the only one that commonly offers the possibility of knowing the physical location of each user. Most of the European carriers have already launched several location-based services.

About friendZone
friendZone is the world’s first location-based wireless community, and was developed by a company called AxisMobile (formerly known as Valis, of which I was a co-founder). friendZone was first launched in Switzerland in May 2001, with the local Swisscom wireless carrier. Four of friendZone’s key services are: location-based dating, enabling users to look for people in their vicinity who match specific profiles, location-based chat, and a friend-finder tool that enables users to physically locate groups of consenting friends. friendZone also serves as the application infrastructure for local wireless brands targeted at the youth market. As of April 2003, friendZone communities exist in Switzerland, Portugal, Germany, Israel, and China.

The location-based privacy puzzle
To better understand the concerns of friendZone’s target users, the design team asked teenagers who had used friendZone for more than a year what they thought about protection, privacy, and tracking services—specifically chat, dating, and the friend finder. The common response we received (accompanied by puzzled looks) was: “Why shouldn’t we feel protected? The only people who know our location are our friends—and we want them to know where we are.” I found this response surprising, since privacy concerns were frequently raised in background interviews conducted with focus groups prior to friendZone’s launch.

Intrigued by this apparent contradiction, I checked the quantitative usage statistics of three carriers in Switzerland, Israel, and Portugal. They indicated that there was almost no use of friendZone’s built-in privacy management tools. These tools offer two main functions: a “disable location” command, and control of a “black list,” which lets users allow (or deny) people the ability to see their location. Total usage of the privacy management features accounted for just 1–5 percent of the total usage of the system. Interestingly, the carrier with the most use of the privacy management tools was also the one in which the system was the least successful. Overall usage of the friendZone system, however, ranked as one of the top three data services offered by the carriers.

Considering the attitudes of the friendZone users and the usage statistics, we can draw one of two conclusions:

  1. Privacy is not an issue, or
  2. In matters concerning privacy, relevant technology and features are only part of the solution.

Conclusion A seems unlikely to be true (common sense tells us privacy must be an issue at some level), so B must be correct. But what is the rest of the solution?

My research and experience suggest that it is possible for users to feel secure even if they don’t actually use the conventional measures of protection (such as disclaimers and blacklists). There is some other element, apart from the traditional privacy measures, that allays the privacy concerns of the users. This invisible link is the psychological aspect of users’ reported sense of security. To best address and fill users’ need for a sense of security, I employed principles of psychology in the design of the friendZone system.

Understanding the target audience
In privacy-related issues, our mission as IAs is to make the user feel safe. We are dealing with the human mind and emotions, and it is therefore imperative that we understand what makes people feel safe, and what is important to them as users. This approach, which operates from a psychological rather than legal or technical perspective, is the essence of PDIA.

Note that recommendation and use of PDIA does not in any way contradict or reject use of legal measures, such as disclaimers, or technological tools. PDIA complements, rather than replaces, traditional legalistic and technological approaches to privacy and other end-user concerns.

In order to better understand how to implement the PDIA approach, we first need to understand who our users are. I will use friendZone as a case study.

The leading location-based services (LBSs) are community tools and entertainment. (See Appendix A for supporting research.) These two fields are generally associated with young users. The term “young user,” however, is a little misleading in this case. The phrase usually denotes an age group, whereas here it refers to a larger public defined instead by its mindset. People in this group are at a time in their life when their world revolves around friends rather than work or family. A 30-year-old single person could fit this description, while an 18-year-old workaholic would not. To retain this distinction, in this article I refer to this group as teenZ (a moniker we coined during the design of friendZone).

What are the characteristics of teenZ in general? Research shows that teenZ are:

  • Constantly changing and unstable.
  • In a process of self definition—transitioning from the world of family and parents to a new, self-defined world of friends and peers.
  • Often in conflict by trying to both be part of a group and establish themselves as individuals.
  • Enigmatic: The attitudes and behaviors of teenZ you meet today might be completely different tomorrow.

At an individual level, a look at the psychological profile of a member of the teenZ group reveals the following needs:

  • Belong to a group
  • Have contact with other members of his peer group
  • Possess status symbols for self-definition
  • Be adventurous–the need to “check out” the world
  • Be in control

(For methodology, see Appendix B.)

A design that can fulfill these needs for teenZ will help them feel more secure and will, in turn, be rewarded with their confidence and loyalty.

TeenZ and privacy
Next, it’s important to understand what teenZ’s concerns are regarding privacy: What are their fears? What do they need to protect? Over what do they need privacy control?

Privacy means different things to teenZ than to other groups. For teenZ, by and large, privacy equals individuality. As teenZ are in the process of defining and forming their own individuality, privacy from parents and parent figures is of the utmost importance. The most essential type of protection teenZ desire is protection from authority. That authority may be parents, teachers, or a boss at work. TeenZ don’t want their parents to know whom they are dating and where they go. In many ways, teenZ try to define themselves in direct opposition to whatever they perceive to be the definition of authority. For teenZ, parents and teachers are the true oppressors. Thus, it is crucial to communicate to teenZ that the “oppressors” will not use the system.

TeenZ also need privacy from their own peer group and security from peer pressure. This need is common among teenZ, but is often neglected. Consider this typical situation: A group of friends wants to do something—for example, go to a certain movie. One member of the group doesn’t want to go, but also doesn’t want to confront the group. What should this member of the group do? If the system can offer a solution or assist in solving such problems, it can instill confidence in teenZ that will make them much less worried about other privacy matters.

Another privacy issue related to peer groups is the need to preserve intimacy. For example, a teen may be dating a new girl, but hasn’t yet decided if it is cool to date her or not, and he doesn’t want his buddies to know about the relationship. At the same time, adding to the complexity of this situation, the teen doesn’t want his friends to know that he is keeping this secret.

Finally, teenZ have the need to protect their image—they must look cool all day long. Thus, they only want their location disclosed when they are at places that are considered “cool.” They’re comfortable letting their friends know they are at a club, but when they are visiting their grandmother, they don’t want anyone to know it.

friendZone implementation of psychology-driven IA
The wireless community itself addresses some of the needs of teenZ by helping users stay in contact with their friends and enabling them to sample, or “check out,” the world and meet new people. In addition, wireless communities are still regarded as a new and cool technology, filling at least partially teenZ’s desire for new technologies and status symbols. Responses to the other needs of teenZ, however, must be designed into the system.

In friendZone, we addressed the relevant psychological factors by incorporating them into the design of three different levels of the system: packaging, tools, and services. Each of these aspects of friendZone utilizes specific subcomponents:

  • Packaging: Language, feel, and alerts
  • Tools: Privacy management tools
  • Services: Traffic generating program and community management

Packaging: language, feel, and alerts
The packaging of friendZone mainly answers the need for protection from authority, the need to be part of a group, and the need for status symbols.

TeenZ-oriented language
Offering teenZ a unique language is beneficial in several ways. First, it helps teenZ feel protected from authority. Having a unique language strengthens teenZ’s sense that their parents will not understand the system. The logical conclusion is that if parents cannot understand the system, they will not use it, thus instilling in the user a sense of protection from authority. (Background interviews in Germany and China indicated that local teenZ prefer English terms because those are less likely to be understood by teenZ’s parents.)“TeenZ tend to have an emotional relationship with their handset. They love it, miss it, dream of it, etc.”

Use of a unique language also promotes a sense of community. TeenZ are likely to feel that if they share specific terms that no one else understands, it contributes to the exclusivity of their community. Finally, from a marketing point of view, creating brand-supporting terms can help raise brand awareness.

When developing a system language to be used by teenZ, I recommend using friendly, comforting terms rather than computer lingo. For example, in friendZone we avoided terms such as “available” or “disable”—we do not normally “disable” our friends, and teenZ might feel uncomfortable doing so. We changed “disable” to “freeze,” and “available” to “free.”

Similarly, it is important not to use authority-related terms—i.e., any words or phrases that might remind teenZ of authority figures in their lives. Try to avoid terms such as “error.” TeenZ already suffer frequent exposure to such terms, usually with a negative connotation, in school and at home. Instead, it may be better to offer a friendly note (“Oops, I think we’ve got a mistake.”) or suggest a solution (“We don’t know the ‘fond’ command, did you mean ‘find?’”).

It is also important to strike a balance between innovative teenZ lingo, and plain, friendly, adult language. Using “cool” terms and cryptic language in excess might end up alienating part of the teenZ audience—some might feel they are dealing with manipulative authority figures in disguise, or just find the system unclear or not user-friendly.

Feeling that the system is your friend
TeenZ tend to have an emotional relationship with their handset. They love it, miss it, dream of it, etc. By taking this into account when designing the friendZone system, we helped make the phone a friend. Remember, teenZ have much fewer privacy concerns toward their friends, so if they regard the system as a friend (to some extent), and not as an authority figure, they will feel more comfortable with it.

One way to provide this friendliness is to design a dialogue with the user to make her feel more comfortable. For example, teenZ don’t want to feel that they are alone or don’t have friends, so in friendZone, all system messages are phrased in a collective, first-person voice (“we” rather than “you”). This helps the user at least subconsciously feel that someone is always there with him or her. Another example of this approach in friendZone is what happens when a user’s name is deleted from a peer’s contact list: the system allies itself with the user. Instead of telling the deleted user, “You were deleted by Sharon,” its message reads: “Don’t bother with Sharon.” (These messages are, of course, adapted to the local culture.)

Segmented marketing campaigns
The marketing campaign for a product aimed at teenZ can also be a tool for protecting them from authority and establishing the product as a status symbol.

One of the cardinal rules of advertising is that the best way to reach teenZ is to distinguish them from both their parents and younger siblings. Key to achieving this distinction is not to market the product in the same way to both teenZ and other target audiences. Technically, a product like friendZone could be sold as part of a fleet management system (to track the location of employees or vehicles) or as a mobile nanny (to track the location of young children), but it would have to be marketed and branded completely separately from the teenZ network. Consumers, and especially teenZ, must never have an inkling that these systems are in any way related.

Tools: privacy management
Privacy management is a basic component of most of the community services offered on friendZone. There are several things worth keeping in mind when designing privacy management tools.

Protection from peer pressure
Protection from peer pressure is implemented in various parts of the friendZone system, but most clearly in the privacy management options. A good example is the command for disabling location awareness (known as “freeze” in friendZone). In friendZone, there is one uniform system notification to inform peers that a certain user’s location cannot be determined. This same notification appears whether the user disables his location, turns off his phone, or even if there is some technical malfunction. This allows users a veil of confidentiality as to why they cannot be reached, and enables them to obscure whether their unreachable status is of their own volition or due to some technical situation. There is also a command that enables the user to disable peers and have those people know that they have been disabled (“X” in the system lingo), which affords a clear distinction between hiding and just being unavailable.

Controlling privacy
An effective privacy management tool for teenZ must address the need for privacy-related information and offer a sense of control. The first can be achieved by providing information regarding the whereabouts of friends, while the second can be achieved by including features that allow users to manage the flow of that information in the system. friendZone test cases showed that 60–80 percent of privacy management usage is spent by users just looking at their black list, without making any further use of privacy management tools.

Statistics show that much of teenZ actual usage of these lists is like “window shopping” to see which peers they approved or disapproved and who asked permission to know their location.

This window-shopping behavior is important for several reasons. All are connected, again, to the need of being in control:

  • It is important for teenZ to know who their friends are. TeenZ want to know both who is on their list and whose lists they are on. TeenZ think that anyone who wants to know their location probably considers them a friend, would like to be one, or has otherwise expressed some level of interest in them.
  • TeenZ always hope that maybe the coolest girl or boy will want to know their location, because such interest could make them more popular among their peers. Thus, it is also important that teenZ have a way of presenting their lists to their friends.
  • Relevant lists can become a kind of popularity index. The person who has the largest number of people on his or her list is, by implication, considered to be the most popular.

Services: traffic-generating programs
Finally, an important and often overlooked aspect of helping users feel their privacy is being protected is the formal management of the system itself. In friendZone, this management was handled by a traffic-generating program (or TGP). While the primary goal of a TGP is to increase usage and adoption of the system, it can also enhance the feeling of being part of a group, which can reduce the various privacy-related fears of teenZ.

The friendZone TGP utilizes several approaches that are relevant to psychology-driven IA:

  • Team management: First and foremost, the TGP is managed by a team of teenZ who are part of the community itself and fit a specific, targeted psychological profile.
  • Feedback: The TGP managers arrange online surveys and conversations with community members, which are intended to give users the sense that if they experience a problem or are bothered by something, there is someone there to listen to them.
  • Hosting: There are always people using the system, so a user will never be alone in a chat room. An effort is also made to have every user be approached at least once a session or a day in the dating section. The concept of hosting is designed to protect the user from ever feeling alone or unpopular.
  • Monitoring: The system tracks and controls deviant behavior to protect users.
  • Newsletters: Users periodically receive a message (via WAP or email) that describes the community’s recent developments and culture. The newsletter strengthens the sense of community, the feeling that “we are part of a unique, cohesive group.” Its aim is also to ultimately strengthen users’ general sense of security.

To summarize this article in a few sentences: Location is important for the wireless community and medium—it enhances the benefits of the wireless medium and contributes to a sense of being in touch and not being alone. At the same time, location introduces significant privacy concerns: It reveals information about us that we may not be accustomed to sharing (such as our exact current location), thus it may be perceived as invasive.

The best way to solve these problems is through use of a psychological approach—by seeking to understand and allay the fears and needs of the users through the design of the system. This approach is psychology-driven information architecture.

This article focused on problems connected to the location-based services, but psychology-driven IA can be applied to any design problem in which privacy is a major concern, as well as other broader information architecture challenges.

Appendix A: Why location is important
Since this paper deals with privacy issues related mainly to location, the first question that should be answered is: Why are location-based services (LBSs) important at all? This question can be approached from two different angles, representing the carriers’ and the end users’ perspectives.

The carrier’s perspective: There is huge revenue potential in location-based services. Analysis Research Consultancy (ARC) predicts that LBS services will account for 40 percent of carrier revenue in 2007. Concise Insight predicts that the LBS market will be worth $7.6 billion by 2009. Analysts at both ARC and research consultancy BCWS predict that the most important location-based services, in terms of usage and revenues, will be in the fields of community and entertainment.

The end users’ perspective: Location information responds to some of the users’ basic needs:

  • Helps the user feel a sense of control
  • Enables the user to know more about his or her peer group
  • Enhances the possibility of being in contact with other people
  • Answers one of the biggest fears of modern times: the fear of being alone. LBS enable users to have information about people and places in their immediate vicinity.

Furthermore, LBS enhances the benefits of the mobile phone in general. The mobile phone affords users the possibility of contacting friends at any time. LBSs take this benefit to the next stage, enabling users to have additional levels of detail about members of their peer group.Appendix B: Psychological methodology
A few words about our psychological methodology: The psychological profiles and insights were developed by Doron Weinstock, the head of the AxisMobile research team. The profiles were first created based on a combination of textbook analysis of teenagers, research conducted by advertising firms (such as the Publicis’s “tween” project), and media companies (such as MTV). This data was combined with results from an ongoing focus group that included eight teenagers chosen to represent teenager archetypes (“the clubber,” “the clueless,” “the idealistic,” “the geek,” etc.). Although all of the teens lived in Israel, they grew up in various counties (Belgium, USA, Russia, Germany, etc.). Additionally, the input received from the ongoing focus group was compared to that gathered from local focus groups conducted prior to each regional launch of friendZone.

The results described in this paper are based on three years of ongoing and local focus group research conducted in Germany, Switzerland, Spain, Portugal, and Israel.

Oded Napchi has been a major force in the Israeli media industry for over ten years. After years of experience in traditional media as the manager of a major radio station, and of the Israeli Children Channel (Israel’s most popular cable network), he played a pivotal role in broadening the horizons of traditional media companies and introducing “new media” concepts and practices. After founding Valis (later re-branded Axis), the world leader in mobile communities, and the first to introduce location-based communities, he is now working as a freelance consultant to television and new-media ventures.


  1. Hi Oded,

    Are you still checking comments on this post? I’m wondering if we could talk more about your work with FriendZone. This piece you wrote is really fascinating.


  2. I have to say I love this article as a case study for how ethnography really makes such a HUGE difference in the design of products and services.

    Oded, I have a question about how you cam up w/ your “persona development”. What were the methods used to get the info.

    I’m also curious as to how you feel your PDIA method differs from Contextual Inquiry or Cooper’s Persona and Goal Directed design methods? Think of this not as a criticism or dismissal, but a request to compare and contrast.

  3. Hi david,
    Thanks for the comment.
    The PDIA is based on a field experience I gathered in the past 7 years while working in different media (tv, internet and cellular) I used it in each one of them. I developed while trying to imply the samba school theory of Seymour Peppert (i hope i spell his name right) of mit media lab. He claims there that community should be built by the community members themselves. The pdia is basically some kind of a “jump start” for this approach or a solution to target audiences who can’t build the community and especially the community tools. Regarding the methodology i explained it in appendix b of the paper; if you need more information please send me a specific question. Regarding the approaches you mentioned, unfortunately i am not familiar enough with them. While working on the paper i got feedbacks that prof. sonia livingstone of LSE has a Contextual Inquiry approach, which resemble in some aspects my approach, is this the case you mention? i mentioned her work in the bibliography of the paper, in any case i will appreciate if you could refer me to the paper you mentioned. i hope i answered your question, if not, please feel free to ask

  4. Hi Oded,

    Ok, I had a printout on the train w/ me when I read the article and the last page must not have printed. I went back and read it again and I see where your methodology appendix is.

    The two references I made in my above comment can be seen in two books by Alan Cooper (most famous for being the inventor of Visual Basic):
    “The Inmates are Running the Asylum” & “About Face”. He writes about Personas & Goal Directed methods. “Contextual Design” is written and promoted by Karen Holtzblatt.

    Both methods though promote more of an ethnographic and less of a focus group approach than you used. I think the researc of sub-culture that you did bears merit. I think purely methodologically speaking that both of them would have used less focus groups and psychological data and spent more time actually with teenZ trying to see how they use their phones. How they make them a part of of their lives in their day to day use and how they then can fill a whole in their service.

    I see PDIA beign useful for the structural analysis but less useful towards innovating the idea in the first place. Both are incredibly important pieces of the pie.

    My next question would be how would the “garage band” version of PDIA work? In order for smaller organizations and groups to get involved in this type of work to prototype it before making it a major piece of their organization they will need to do it on a discount basis w/ shorter time frames. How might you see that happening?

    Lastly, if you were to do this project now for another age group, like Yuppies, how might you do it differently? What did you learn in your results through this process that would make you ammend it slightly?

    As you can tell, I’m incredibly interested in this method. I have a strong background in culture & personality and in UI design and combining the two is something important and useful to me. I also really appreciated the real-worldness of your example and this is not just a graduate student study, but a real application of a methodolgy with measured results.

  5. Oded,
    At first glance, I thought it was unusual the way you use the term “psychology” in the title of your approach. I was assuming the primary and most well-known definition of “psychology” – “the science of mind and behavior” – and was waiting to see how your use of the term was relevant. On second read, I realized you must be going with a secondary definition of “psychology” – “2 a : the mental or behavioral characteristics of an individual or group”. Is the term often used this way in your line of work? The word encompasses so many things and it seems like you are referring specifically to perception of or feelings for a product/service. I’d love some clarification on this if you (or anyone else) are willing to offer.
    Many thanks.

  6. Hi David,
    First, sorry for the delay in my answer. I will answer just part of your question now, and the rest will follow later.

    I used the pdia method in the past also in smaller scales and for other target audiences. The main points I try to keep in small scale are:
    Psychology analysis of the target audience. Try to sit in brainstorming with a psychologist who is familiar with the target audience or, if it is a company, you can do it with VP HR or Organizational Behavior consul. Try to understand from them what are the needs of your target audience? What are they fears? Try to embed this insight in your program (even small things as a possibility of putting your daughter picture as a wall paper to CRM application). Use this brainstorm also for building archetypes of people in the target audiance

    The second is find a group of people who give a good representation of your target audience, using the archetypes you developed in the previous point. Meet with them, think with them and understand what they need, I believe that 1 session of one on one and 2-3 sessions with the group could give you some excellent insights.

    This basic usage of the pdia should take around a week of work, I found that in most cases I can spare a week for designing the system.

    Part 2 in few days…

  7. Dear Katie,
    First, thank you for the response.
    I am not sure I understood your question, but if I got it right, it seems as it was taken from the semantics field (i.e linguistic philosophy).
    If so, I think that if you are trying to get to full meaning of a term the explanatory (based on B. Russel and G. Frege) approach you used is limited and therefore the understanding of the term is not complete. I prefer the approach of understanding the meaning through the use, suggested by L. Wittgenstein. (this is the basic conflict between B. Russel approach and the later L. Wittgenstein). So if I will follow the wittgenstein approach I will claim that in order to understand the meaning of the term “psychology” you should look at the work of psychologists. This the meaning I refer to in the term psychology. But if in any case I should use explanation, I will prefer the britanica’s one: “scientific discipline that studies mental processes and behavior in humans and other animals.”, since the driver of my design the possibility to analyse the behavior pattern of the target audience.

    Hope It answered you question, if not please e.mail me again


  8. Hi Oded,
    Let me start by praising you article and thanking you for sharing your insights with us. The psychological facet is undoubtedly an important part of all any AI effort and you have elucidated the importance and methodology for creating in beautifully. However, I feel there is something, which needs further clarification.

    In “the real world” (whatever that may mean) “…seeking to understand and allay the fears and needs of the users…” will not provide you with the solution to the much nastier issue of addressing the fears and needs of “the system”. It is exactly the forces outside the community of users, those who believe the needs and especially fears of the users are not nearly substantial enough (“they don’t know what’s good for them”) which will coerce you into compromising the PDIA you developed.

    True, PDIA will be used to soften the blow of those imposing external forces (using your excellent example – by creating a language which will be less intimidating), but if the system is successful these forces will eventually weaken the psychological tools to such an extent they will become mere frills.

    If my hypothesis is correct, then the case you referred to as “…the carrier with the most use of the privacy management tools was also the one in which the system was the least successful…” was the one, which imposed most restrictions.

    Then again, I could be dead wrong….

  9. This is a very well written article. You should write more. Waiting for your next piece.

Comments are closed.