Asking Why: A Conversation with Faith Bolliger

efcf0a9b-20ed-43d9-91a1-7e8e02923886Faith Bolliger is the Head of Service Design at AKQA, a global design agency with offices in North and South America, Europe, and Asia. Faith began her career in UX fifteen years ago and, over the years, she has held a variety of in-house and agency positions in both New York City and San Francisco. Faith was kind enough to speak with me recently about her experiences working in UX, the skills she feels young UX designers should have, and some of the opportunities and challenges facing the UX community today. I chose to speak with Faith because of the variety of her professional UX experiences.

You hold a bachelor’s degree in Sociology from the University of Illinois at Urbana-Champaign and a master’s degree in Sociology & Historical Studies from the New School. Do you feel your interest in sociology is apparent in the ways you approach UX design?

Sociology is concerned with the questions of “why?” Why do we have racism, sexism? Why do we have poverty? The why questions, which attracted me to sociology, are one of my chief interests as a UX professional. There are a lot of people in the profession who focus on the “how.” How do we design something that is innovative? How do we design something that is unique? But for me, questions of how have always been secondary. I like to ask why? Why are we making the design choices we’re making? Why will our solution deliver value? Why will people want to use it?

Your first position as a UX Designer was at McKinsey & Co., where you started as a part-time temp. You’re now the Head of Service design at AKQA. Could you talk a bit about this unique career trajectory?

My introduction to UX was serendipitous. I began working at McKinsey part-time as a temp, in 2001, while I was completing my graduate degree. After I completed the degree, I decided not to pursue a PhD and focused on going to law school, instead. I felt that the law would allow me to apply what I had learned during my studies in sociology and have some impact. While I was waiting to attend law school, I continued temping at McKinsey and was eventually asked to interview for a generalist team position in their Internet Communications department. I was hired and had to choose a focus among development, project management, or design. I chose design because I knew a lot of people working in design at the time and figured it would be fun to learn some of the programs, such as Photoshop and Illustrator. I ended up not going to law school, and instead, began taking a variety of night classes in design, color theory, and illustration, at SVA, Parsons, and NYU. I wanted to strengthen my design sensibilities and these classes helped me to achieve this goal.

A couple of years later, I transitioned to information architecture during what was proving to be a particularly exciting time: technology was beginning to influence what we could do with the browser and there were many more opportunities for in-page interactions. Applications and software were also being released that users could customize. Everything was becoming more interactive. From that point, I really started getting into user-centered design, which meant advocating for user testing.

After I left McKinsey and moved to San Francisco in 2008, I took a job at the brilliant interaction and design services consultancy Cooper. Their design and research methodologies are heavily interview- and ethnography-based, so it was a great fit for someone with my background. In my current positon as Head of Service Design at AKQA, I no longer work on interactions within mobile products or websites; I’m looking at cross-channel interactions, touch points…the entire ecosystem.

While you began your UX career as an in-house designer at McKinsey, the majority of your work since then has been at agencies. Was the transition from in-house to agency difficult?

Even though I was technically in-house at McKinsey, our team acted more as an internal consultancy. We had a variety of in-house clients and we built products for those clients. For example, we built time and expense reporting tools for the HR and Administration departments. I feel like at McKinsey I learned how to be a small agency. I still had stakeholders who had to make the final decisions and I still didn’t own the product, which was very similar to how I worked at the agency Cooper. At AKQA it’s been more of a mix. Some of the accounts I work on are on retainer and that means we have an ongoing relationship that gets renewed every few years.

As the manager of a UX team at an agency, what are some of the qualities you value most in the designers working under you?

It’s critical that my designers be ambitious and take initiative. Designers who choose to work agency-side need to have these qualities. Agencies can be a great place to learn design, because of the variety of work being done there, but they are not environments well suited to hand-holding. If you really boil it down, clients are paying for work and that can create a lot of pressure; constant pressure, however, can bring out the best work in some people. Still, many designers who have decided to work agency-side are finding themselves and their agencies being acquired by companies with deep pockets who want to incorporate them into their in-house teams. Hot Studio’s acquisition by Facebook and Fjord’s acquisition by Accenture are just two examples of this ongoing trend.

Learning skills on the job is a common expectation for many professionals. What types of skills should new UX professionals be expecting to learn on the job?

During the first few years, it’s just about learning all of the tool sets and how to structure your design thinking. When starting out, it may feel very tense when you’re uncertain of how your work will be received, but people are expected to hit some road bumps and have some problems along the way, so you shouldn’t take the judgments too critically. It’s just normal growth. Learning how to interact with clients and learning how to manage teams are skills that are difficult and stressful to learn on the job, but unfortunately, there’s no other way to learn them. Handling a conversation with a client that’s gone wrong, particularly when you’re new to working directly with clients, can be very stressful, as can leading a team of designers for the first time.

We are currently beginning to work with prototyping tools for our final assignments. Do you have a favorite prototyping tool you’d recommend to designers new to UX?

My favorite is still old-school paper prototyping. It’s just easier. It’s also great because you can work through a lot of ideas and shop them around very quickly. When you start getting digital, it’s slower and cumbersome, and you become really fixed to it—you’re a little less open creatively to changing ideas because you have already spent so long designing it and getting it built.

Are there any emerging trends in user experience that you’re particularly excited about?

I know that some people in the industry think VR (virtual reality) is going to blow-up. In terms of the industries I’m mainly working with, which are retail, e-commerce, media, etc., improving upon cross-channel experiences and the tools that support these experiences are areas I’m excited about. Taking your online shopping experience into the store, or store experience back online…that’s something that hasn’t been very smooth. I’m interested in closing that loop; of bringing someone who hasn’t bought your product into your brand, getting them to purchase, and then retaining them.

Many of the students in my classes will be graduating at the end of this semester. What kind of a job market do you think they will be finding themselves in?

There’s a lot of opportunity out there for both agency and in-house positions, at least in the San Francisco area, which I’m more familiar with these days. Universities have just recently begun understanding how to teach UX, so the field is by no means saturated. There seems to be much more of a demand for UX designers than supply.


Retail’s Getting Details: New Data Collection Tools Offer Retailers Insights into the Usability of Their Brick-And-Mortar Stores


source: DADI+

Usability testing, fundamentally, seeks to understand the ease with which a user can find what he or she is looking for in a particular environment. The retail industry has always had a unique interest in understanding the usability of its environments as customer confusion and frustration can quickly translate to lost sales. While a variety of tools have been available to retailers who wish to test and improve upon the usability of their online environments, retailers who would like to better understand the usability of their offline (brick-and mortar) environments have had very limited options. Fortunately, recent technological advancements have changed that.

Currently there are numerous companies offering tools that can be employed by retailers to test the usability of their offline environments. One such company is Euclid Analytics. Euclid was founded in 2010 and is currently contracted by over 500 brands operating in 45 countries worldwide (Euclid Analytics, n.d). Euclid offers retailers the ability to gather a variety of information about the customers visiting their offline environments, including the length of customer visits, where the customer goes within the store, how much time the customer spends in each location, and how frequently the customer returns to the store. Euclid’s data collection tools are unique because they do not require retailers to install expensive analytic-specific hardware in their stores, nor do they require customers to install retailer-specific apps on their smartphones; instead, Euclid collects customer data by employing the retailer’s Wi-Fi connection and linking that connection to the customer’s smartphone (Shu, 2016).

Customer information, once collected, is anonymized, aggregated, and displayed using the Euclid dashboard. This dashboard incorporates a variety of data visualization tools, which allow retailers the opportunity to analyze their data and make changes to their stores in order to improve sales (Datoo, 2013). Retailers could, for example, improve store pathing by adding additional displays to areas of their stores that receive the least amount of traffic or allocate additional staff to the sales floor when their stores are statistically the busiest.

Euclid’s data collection tools have two obvious drawbacks, however: First, because Euclid relies on connectivity between the retailer’s Wi-Fi and the customer’s smartphone, Euclid cannot collect data from customers without smartphones or customers who have their smartphone’s Wi-Fi function turned off while shopping in the store. Second, customers are likely to have privacy concerns with regard to the data collection process, despite Euclid’s assertion that all data collected is anonymized. Nordstrom discontinued its use of Euclid’s data collection tools after customers voiced dissatisfaction over being tracked while shopping (Clifford & Hardy, 2013). If customers feel that the collection of their in-store data is an invasion of their privacy, then the data collected, regardless of its value in improving a store’s usability, will, on the whole, hurt the retailer more than help.

Fortunately, there are ways retailers can maintain customer trust while collecting useful information that will help improve their stores’ usability. Retailers need to be up front about the types of data they are collecting, how they are collecting the data, and how the data will be used. Retailers should also ask customers to opt in rather than require them to opt out of having their data collected while they are in-store (Datoo, 2013). Incentivizing the opt-in process by providing special offers and discounts to customers who agree to have their data collected can improve the size of retailers’ potential data sets.

Analyzing the usability of offline environments seems to be the next frontier of usability testing and I, for one, am particularly interested to see how things unfold. In order to test these environments successfully, however, companies such as Euclid Analytics and the retailers who employ their data collection tools must be sensitive to the fact that customers who may be comfortable (or at the very least, resigned) to having some of their personal information collected online may react very differently to having comparable information collected in person, even if it is, in fact, anonymized.



Clifford, S. & Hardy, Q. (2013, July 14). Attention, shoppers: store is tracking your cell. The New York Times. Retrieved from

Datoo, S. (2013, October 11). What information can retailers see when they track customer movements? The Guardian. Retrieved from

Euclid Analytics. (n.d.). Retrieved from

Shu, C. (2016, January 14). Euclid analytics raises $20m series c to track consumer behavior in retail stores. TechCrunch. Retrieved from

[header image via DADI+ (]

Design Critique: Lusmo® Automatic Pet Feeder



The Lusmo® is a pet feeder of Japanese design that automatically dispenses food for small dogs or cats and can be programmed to provide up to three meals per day. Pet owners can adjust the quantity of food dispensed (from 1/16 cup to 1 cup) and the time of each meal. I’ve been using the unit in my home for two months and am very happy with it; it is reliable and beautifully designed. Programming the unit, however, is a complicated process that could easily be improved with the implementation of a few simple changes.

Fig. 1: The Lusmo®
Automatic Pet Feeder
(Model L-AF120R).


Improvement Opportunity 1

All programming tasks are completed by engaging the feeder’s three multi-function programming buttons in a variety of sequences and combinations.

Recommended Design Change 1

Simplify programming tasks by giving each task an independent button with corresponding label.


Fig. 2: Diagram of the feeder’s display panel, which can be found in the feeder’s
instruction manual.

The feeder’s display includes a digital screen and three programming buttons (two are represented by the black triangles to the left of the digital screen; one is represented by the black circle to the right of the digital screen). Each programming button performs a variety of functions, depending on the following:

  1. How long the button is pressed.
  2. At which point in a sequence the button is pressed.
  3. Which button (if any) is pressed in combination with it.

This design minimizes the number of programming buttons required to operate the feeder but greatly complicates the feeder’s system mappings, as there is not a logical one-to-one relationship between action and outcome.


Fig. 3: Instructions for programming feeding times, which can be found in the
feeder’s instruction manual.

The feeder’s complicated system mappings are particularly apparent when programming feed times. In order to set a feed time, the user must complete steps 15 listed in figure 3. This sequence is arbitrary and the labeling on the unit itself does not provide adequate instruction nor direction. Thus, in order to change the feeding time, the user must either,

  1. memorize the sequence (what Norman refers to as “knowledge in the head”), or,
  2. have the instruction manual readily available whenever the feeding times needed to be set or changed (what Norman refers to as “knowledge in the world”).

In this instance, neither option is ideal. Illogical sequences such as this, are first, difficult to remember, and second, difficult to retrieve. Using “knowledge in the world” is a much better solution when approaching arbitrary sequences; however, this type of knowledge is only useful if it is available at the moment the sequence needs to be initiated, which, in this case, would require the user to consult the instruction manual whenever the feeding times needed to be set or changed.

Compounding the aforementioned issues is the fact that the sequence must be done quickly: if ten seconds pass between any of the steps in the sequence, the user must begin the process again.

There is also a much more problematic issue with mapping all of the feeder’s functionality to three multi-function programming buttons: namely, that one of the programming buttons used in a variety of sequences also turns off the feeder. If the programming button that controls the power to the feeder is pressed long enough, at the wrong moment in a sequence, the unit could be turned off. This in turn would leave the pet without food for as long as it would take the user to discover the error.

This outcome is unlikely, but the mere possibility of such a serious mistake occurring indicates that the system design does not properly take into account user error.


Fig. 4: Proposed redesign of programming buttons. The power button is offset from
the other programming buttons and given a unique shape in order to reduce the
likelihood of the feeder being turned off by mistake. 

By reorganizing the display panel to include separate buttons for Power, Setting, Mode, Manual, Forward, Backward, Increase, Decrease, and Enter (and labeling each button accordingly), logical sequences for programming the feeder could be developed and implemented, which would greatly improve the ease in which the feeder could be programmed. The redesign would also greatly reduce the risk of the unit being turned off by mistake.

Improvement Opportunity 2

When programming the amount of food to be dispensed by the feeder, the user must select a code, which corresponds to a particular quantity of food, rather than the quantity itself.

Recommended Design Change 2

Replace the current digital screen, which is only capable of displaying a four-digit alphanumeric string, with another, more robust digital screen that would allow for longer alphanumeric strings. Then, the actual quantity of food to be dispensed could be displayed and selected.


Fig. 5: Instructions for setting the feed amount, which can be found in the
feeder's instruction manual.

The process for programming the amount of food to be dispensed employs the same three multi-functional program buttons as the process for programming the feeding time, so implementing Recommended Design Change 1 would also be beneficial to this process.

What is unique to this process is how the necessary information is displayed on the digital screen. Instead of selecting a particular quantity of food, the user selects an “F” code, which corresponds to an actual quantity of food (see diagram in figure 5). While the pairings of the codes to the food quantities are logical (i.e., larger codes=more food), they still require that the user either,

  1. memorize the code-to-quantity pairings if the user would like to select a specific quantity (“knowledge in the head”), or,
  2. have the instruction manual readily available whenever the feeding quantities needed to be set or changed (“knowledge in the world”).


Fig. 6: Proposed redesign of digital screen. The screen would be capable of
displaying a complete feeding quantity with unit of measurement.

Again, neither of these options is ideal. If the digital screen were capable of simply displaying a few more characters (as seen above), the “F” codes would be unnecessary because the quantity of food to be dispensed could be displayed and selected instead.