This week we read two papers (link1, link2) about interacting with visualizations. They outlined the different ways that individuals select, filter, navigate, re-organize, and store information through interactions with visualization systems. For this activity you will work together in a small group to design a visualization-based dashboard that features interactions. Think carefully about how users will spot trends, filter data down to fit their interests, see an overview or details, select points of interest, handle millions of points onscreen, and share their findings. The goal of this activity is not to produce a final, polished design. Instead, aim to create a general "sketch" of your visualization. You should be able to tell the class how you decided to visualize the data (ideally with a rudimentary sketch), the specific interactions you have brainstormed implementing, what choices you made during the design process, and any trade-offs or alternatives you considered. In this activity every group will work on the same challenge.
Please follow this procedure:
Your group is an elite team of UX designers at an up-and-coming Internet commerce firm, Nile. Nile sells a variety of products globally through a hugely popular online platform. They also allow independent sellers to create their own "Nile Storefront" through which they can market their own items on the platform. The products sold are incredibly numerous, and the number of sales each year on the platform reach into the billions. Some independent sellers sell only a few thousand products each year, while others sell millions. One major challenge for Nile's continuing growth is its aging internal infrastructure. Staff depend on a dated internal dashboard that allows them to explore global sales data. They use this tool to identify upcoming sales trends, new markets, areas where they are falling short, and demographics for advertisements (among other tasks). Independent sellers also have access to a version of this dashboard that shows only their own store's sales. The tool is critical to Nile's success as a business.
As is the case with many rapidly growing technology companies, their infrastructure has not kept pace with their growth. The current dashboard, developed in 2003 and only updated in a minor way, uses hyperlinks and requires users to type their own queries in SQL in order to find data. Visualizations are rudimentary, and the JPEG-based map view features a few countries that no longer exist and clickable arrows for navigation. Your team has been tasked with a complete re-write of this dashboard. You are to create a new, highly interactive tool for exploring these data. While you will have a couple of years to complete a full tool, you must deliver a proposal outlining your visual and interactive elements as soon as possible.
Here are some example use cases to consider (but feel free to make up your own):** The "raw" data you can access are a table of individual sales made on Nile. Employees can access all of the data, while Storefront owners can only see their sales. Assume that they have somehow resolved any internationalization / language issues.
Name of item sold | Descriptive title of item that was sold |
Item categories | Tags/categories describing item (e.g. "camera", "housewares", "60w lightbulb") |
Sale timestamp | Time that the item was purchased |
Sale quantity | Amount of item that was purchased |
Purchase price | Cost of item |
Total purchase price | Cost * quantity |
Average product rating | Review rating of product at time of purchase |
Buyer's review | If buyer reviewed, score and review |
Location of sale | Lat/lng coordinates of the buyer |
Geo-info for sale | City, State, Country for the buyer |
Time it took to arrive | Time in days between order and delivery |
Anonymized buyer ID | Identifier for the buyer |
Buyer demographics | Hyper-detailed demographics (e.g. gender identity, age, relationship status, region of birth, interests, possessions, bad habits) |
** Your database team at Nile is very smart, and they are happy to put together some aggregated data to make your life easier. Here are a few examples:
Average sales over a week for a set of states/countries |
Total monthly sales for specific products by demographics - Do all cat owners buy litter online? |
Highest rated product bought by at least 10 people in Florida every March |
Standard deviation in delivery time for a specific city during the last week |
**You've also got access to an internal AI platform that tries to predict future trends. Here are a few examples of what it can provide:
Expected sales next week for a specific region |
Whether sales in a region matched algorithmic expectations last month |
The likelihood that a specific buyer demographic will buy a specific product |