Results
-
4 Interviews
-
2 Personas
Background
I received a take-home design challenge from a government contractor I was interviewing with. The challenge was open ended and they said I could work on any site, and told me not to limit myself to government websites. However, I really wanted to refresh the site of the U.S. Bureau of Labor Statistics site for the Consumer Price Index, a measure for inflation. The challenge had a relatively quick turnaround and thus, the fidelity was only required to go to wire frames.
I’ve been a fan of NPR’s planet money for a while and one of the main topics they discuss has to do with the Consumer Price Index (CPI), which is defined as a measure of the average change over time in the prices paid by urban consumers for a market basket of consumer goods and services.
Essentially, the BLS looks at the price of goods, such as a pound of bacon, a gallon of gas, etc. and maps the price changes over time to gauge inflation at the consumer level.
Why I chose the CPI
-
Navigation
There were too many avenues on the CPI site for the average user (like myself). Because of this, I hypothesized that other users would struggle as well.
-
General Confusion
There is an interactive graph on the page, however it only shows the 12 month percentage change and doesn’t offer much in terms of customization
After choosing my problem set, I made two assumptions based on my own interactions with the site.
From those assumptions, I generated a few user questions.
I wanted the interviews I conducted to be usability/heuristic focused so that I could gain a better understanding of how new users to the site would use the page.
Assumptions and questions after a quick brainstorming session focused on the problem set of not being able to find the price of goods.
In total, I conducted four user interviews
“I feel like there were too many words”
After the four user interviews, I boiled down the responses to 2 user pains and the solutions they personally took to solve the problem
Shattering my Hypothesis
One thing I was not accounting for in my hypothesis, was users utilizing the site’s search function. Despite using this, and the site returning the proper results and data, users were still confused…
-
Poor Search Results
When users searched for bacon on the site, the first result the BLS search function kicked back was a link that had what appeared to be a BLS data identifier that users most likely wouldn’t know correlates totheir search parameters
-
The Right Answer
The First link actually provided the data that users were looking for! they just had no idea because it wasn’t indicated in the results.
-
An Easy Fix
The easiest and leanest fix to this issue would be to simply change the first link to match the results as they would be understood by a human.
Too many choices
Continuing on the homepage’s confusion, I wanted to dive into why users were struggling on the homepage. After some research I found a heuristic principle that defined the issue on lawsofux.com. According to the site, Miller’s law states that the average person can only keep 7 (plus or minus 2) items in their working memory. Therefore, on the homepage, the average user had too many choices on the homepage.
After I conducted user interviews, I synthesized two personas with different needs and demographic data.
Cas U. Aluser (a casual user) was someone who’s demographics, needs, and pains mimicked my own, as well as those of the users I interviewed. Cas is representative of a non-power user. An average browser who’s curious about the CPI
On the other side of the spectrum I developed E.C. Onomist (an economist), a power-user persona who knows the CPI site thoroughly and uses the data on the site to make large financial decisions. For this speculative analysis I did not design for Mr. Onomist.
How Might We…
I decided to keep going with the project despite the easiest fix being the link, as I felt the homepage needed a refresh, so I did a How Might We session to come up with potential fixes to the site’s problems
One of the quickest design studios one can do to generate a ton of ideas in a relatively short amount of time is Crazy 8’s. So I grabbed a sharpie, and in the parking lot of a Pho restaurant I got to work.
I decided to go with a dashboard approach to easily show data immediately. My overall idea was to give users the option to compare the percentage changes of multiple variables so they could see the change over time. In a sense, I aimed to make the CPI a line chart like stock readings.
During the interview, my work was questioned by the design team. Some feedback I got was that maybe a dashboard wasn’t the right way to present the data. The problem is, since this was speculative work, getting access to users of the CPI site would be incredibly difficult. So I decided to create another mockup. In this mockup, the user is overwhelmed by ALL the choice, they don’t want a graph, they just want to search for the results in the CPI database.
Suppose this user wanted more information on the CPI. After the interview, I wireframed a potential CPI homepage based off of various designs for informational homepages from the Baymard Institute.
I wish I had power users to interview, but unfortunately I don’t know any economists so it was difficult finding power users to interview. That being said, if a public site isn’t optimized for the common user, it probably isn’t going to perform very well.
Key Takeaways:
Can it be even simpler?
Research and validation is always good
This was a really fun exercise and the team that interviewed me thought my presentation and methodologies were quite thorough, which definitely put me at ease. My timeline didn’t exactly line up with the company’s openings, but I’m grateful for the opportunity to interview with them and the level of professionalism from each of their design team, even when they were grilling my designs ;)