Designing a crosstab table configuration experience
Role: Lead Product Designer for Analyze
The compare feature in Analyze was quite limiting in that it only let’s you compare two questions at a time and you can only have one compare rule on at a time.
I set out to design for a more advanced comparison feature, a crosstab table which looks for statistical significance or relationships between multiple variables such as age, gender, employment status, and the things these variables may influence such as health, purchases, or travel for example. And where observed differences that are statistically significant are typically highlighted to help you find meaningful correlations in your data.
It was exciting to add crosstabs to the reporting and presentation flow that I had designed earlier thus giving me the opportunity to scale the original designs for multiple reporting assets in addition to the results dashboard and provide an easy way for our customers to share and collaborate.
I designed primarily for more advanced users who are most likely building crosstab tables outside of the SurveyMonkey product. In keeping with our company wide personas, Project Paula and Analytical Anna, I developed the proto-persona, Peter, a Market Insights Manager, to sustain empathy for our crosstabs customer.
Peter, Insights Manager
Peter is a data analyst with a niche in market research where he applies his intellectual and analytical skills to deliver insights to his clients and stakeholders.
- Agency model: Conducts projects end to end, exports the report, packages and shares it out to stakeholders
- Templated task: Peter already does this in Excel with a pre-formatted template and has a pretty good idea up front how to assign data to the banners (columns) and stubs (rows)
- Presentation mode: Once Peter is done exploring, he’ll move into presentation mode ie he’ll prepare the crosstab report to share out with others
Interactions & user testing
The experience of building and configuring a crosstab table was a main focus along with scaling the reporting and publishing flow I had already designed for results dashboards to accomodate for another presentation asset.
Key areas of design included:
- access points to discover crosstabs
- adding question data to a row or column
- editing the question answer options
- re-ordering row and column elements
- swapping rows and columns
- viewing statistical significance within the table
- creating a crosstab report, publishing and sharing it with collaborators
- managing (editing, viewing, deleting, etc) your crosstabs
I took these key design areas along with my assumptions into a round of user testing in order to gather feedback and evidence to inform subsequent design iterations.
After our initial release we’ve been monitoring adoption and are already gearing up to enhance the crosstab table by adding more values to the row and column cells. In addition, we’ll work to improve discoverability, and we aim to add more types of data that can be added to and configured into your crosstab table.