As a UX designer, I take the science and art of ethnography seriously. This practice takes people watching to another level. Ethnography helps me better understand the actions of individuals and groups. Why do people do things a certain way? Why do people avoid certain things? What I’ve discovered over the years in UX design is that even the minutiae of details add up to a larger picture. This snowballing impact of human behavior can be for better or for worse.
A few years ago, HYPR worked with a leading US consumer bank to provide a security solution for a specific need across their branches. I had the opportunity to perform a deep dive assessment of their needs by visiting a branch in Manhattan. The visit was structured — we knew the types of data we were required to capture to better understand the pain points of the branch employees. That way would be able to figure out how to join them in addressing their challenges.
The plan was simple. My goal was to:
- Observe what employees of various levels do and how
- Interview employees and branch management
- Ask employees to perform specific tasks and observe the outcome.
After speaking with employees and branch management, and observing their interactions with customers, I noticed an interesting pattern.The branch tellers appeared to have a methodical system for interacting with the customers. I asked the management team if I could observe the tellers from inside their work area — from their point of view. With managerial accompaniment I was granted permission to enter the teller’s area for closer and more granular observation.
I requested the manager to protect my anonymity and purpose there, so as to not affect the tellers’ behavior. I stood near the tellers and watched them go on with their tasks. Within 10-15 minutes of observing their routine I realized I should time their various actions. There is, of course, significance to the amount of time spent on each action as they engage with customers and technology.
I noticed a triangle of interaction that each teller managed. They had to move between the customer, computer, and printer. As I observed the teller pivoting among the three activities, I discovered an order to the interaction:
- Customer comes to the window
- Teller politely greets the customer
- Teller asks the customer what they can help them with
- Customer replies with a request
- Teller complies and asks for identification
- Customer provides identification
- Teller thanks the customer
- Teller processes the customer request
- Teller locks their workstation
- Teller walks to the printer
- Teller obtains the printed document
- Teller returns to workstation
- Teller unlocks their workstation
- Teller makes sure everything is in order and that the customer request is successful
- Teller hands the printed paper to customer
- Customer thanks teller
- Teller replies with “You are welcome”
- Customer turns around and leaves
Now, believe it or not, this entire flow of interaction took only 2–3 minutes for an experienced teller — very impressive!
It’s important to note that there are two steps the teller does every time, and that is locking and unlocking their computer. People don’t realize the amount of time spent on typing a password to unlock a computer. Something as simple as unlocking a computer seems to not take up much time. We do it all the time because it is a repetitive function for us, and it develops into a habit.
What I discovered was that a teller spent on average 2 seconds to hit the CTRL+ALT+DEL keys plus another 3 seconds to type their password. That added up at least 5 seconds per unlock, that is if the password was correctly inputted the first time. Now, 5 seconds may seem fast and have no meaningful impact, but when you calculate how many times a teller unlocks their computer during an eight-hour shift, in the long run it makes a huge impact.
These numbers do not factor in the additional amount of time required when mistakes are made while typing in a password or missing the CTRL+ALT+DEL keys. I observed and calculated the following:
- Each successful unlock averages 5 seconds
- A teller unlocked their computer about 100 times during a full-time shift
- This particular bank had approximately 100,000 tellers across all their branches
- 5 sec x 100 times x 100,000 tellers = 50,000,000 seconds
That’s approximately 14,000 hours per day spent on typing passwords — talk about productivity losses! If we base this off a 261-day work year, and an average teller pay of $15 per hour, it amounts to nearly $55 million for the entire year. And, this is just at one bank. There are approximately half a million tellers in the US alone, and countless more globally.
The study reinforced that passwords are a massive problem to be solved, and not only in the financial services industry. As a UX designer, bringing passwordless authentication to the world is an exciting challenge. Had I not joined HYPR, visited the bank branch, and embedded myself into the tellers’ environment, I would not have discovered the considerable amount of time we lose due to using passwords.
The takeaway here is that unlike other means of studies and research, ethnographic research is by far the most valuable one when it comes to truly understanding your users. Meaningful UX requires research, interviews, usability tests, and so forth. Nothing can compare to watching your users or potential users in their natural environment. Close observation will likely yield unexpected and valuable results that will help inform how you design for people.