The New York Times created an MVP of an internal ad insights tool called Kaleidoscope to provide data-driven marketing strategy insights to brands purchasing ad space on the paper. I was tasked with creating a new design framework for extending the product to add two additional data dashboards as well as revamped UX and visuals for the result pages.
Interviewing users of the platform as well as reviewing user surveys, I found that there was a need for a more simplified and robust search function, more contextual information for results, and ready-made visualizations for screen-sharing. After auditing all the dashboards I was able to determine whether each needed its own dedicated space or could be more unified within the Kaleidoscope platform.
Taking a highly collaborative approach, I involved focus groups of users as well as the product team at the New York Times to validate my design concepts. A big part of my approach was to utilize low-mid fidelity wireframes and prototypes and work iteratively to get buy-in from all key stakeholders.
The search navigation within the MVP of Kaleidoscope required users to interpret the meaning of a mad-lib statement involving multiple dropdown inputs. As a result, the input and output of the results were not clear to users and required more cognitive load.
The previous homepage for Kaleidoscope didn’t explain or educate visually about what the tool could do for users. For the redesign, explainer text and trending insights were added so users could quickly understand the platform.
To simplify the search experience, the new designs featured one search input where users could limit their search to "editorial engagement keywords" like "audience segments" and areas of the New York Times paper like "sections" and "topics" or "ad engagement keywords" such as "ad formats" or "ad categories".
A mega-menu with exposed keyword lists aimed to help users understand all their available options as well as limit confusion around what users would be searching for. Subtext helped set expectations on what the output of their search input would be.
User interviews revealed the need for being able to search and make comparisons across multiple ranked lists as well as have ready-made visualizations. In the previous version of Kaleidoscope, the results pages relied heavily on explainer text called talking points to summarize numerically ranked results. Talking points were usually copied by users for use in client presentations to help explain the meaning of the rankings but often did not provide quick visual takeaways from the results.
For the new results pages, bar charts were used to visualize the difference and importance of each ranked search result. In addition, users could customize their search result page to view only the categories of information of interest to them. This added flexibility to the results while retaining the underlying interaction paradigm.
The talking points were kept as a feature as it was frequently used to support the recommendations made for marketing campaigns. However, instead of making them the primary emphasis in the UI, they were placed within a tooltip for users who needed help interpreting results and still wanted to copy a “talking point” for their presentation.
It was clear that there would need to be another area to fit all of the additional contextual information needed to be accessed within the search results. This contextual information included descriptions, data collection methodology, demographic information such as age and gender, trending articles, and newsletters. To address this need a contextual side panel with separate content tabs was added. This would allow for any amount of context to be placed and surfaced when needed.
It was important for the Kaleidoscope team to get buy-in and increased engagement with the current users of the platform. As a result of the redesign, we saw an increase in the number of users of the platform, the use of advanced features such as historical trends and comparing categories, as well as an increase in downloading insights.
Something that stood out was remembering how important engaging with the engineering team early on at the Times was. As designers, we need to understand the limitations of the underlying technologies we are working with. This helps scope our work and set realistic solutions. Due to the collaborative and early involvement of all team members and users we were able to design something truly useful and functional. In the future, I’d like to conduct further useability tests on the revamped designs and optimize the experience for users having trouble navigating the redesign.
Demystifying data collection education at the Pacific Science Center
All rights reserved
