USABILITY TESTING
Car Insurance Rates Widget
From existing user behavior data, we know ValuePenguin insurance users are looking for rates.
I proposed creating a new tool where users can select their state to see the cheapest rates based on coverage type.
Problem
Current method of adding large datasets often leads to errors. Writers and analysts have shared that adding data heavy tables in the CMS is time consuming and often leads to user error, leading to tables with incorrect data and differing UI across pages. The SEO team want an easier way to include a rates table with data from a single source of truth.
Poor engagement. Existing tables are engaging users but the cause of dead clicks.
Testing prototypes
Desktop Prototype A
Desktop Prototype B
Usability Testing
We screened for people who are actively looking to switching to a new car insurance company.
-
What is your initial reaction to first table? [Verbal response]
What elements do you think are clickable? What would clicking on it do? [Verbal response]
What is your initial reaction to this second table? [Verbal response]
What elements do you think are clickable? What would clicking on it do? [Verbal response]
Which version to you prefer? [Multiple choice: First one, Second one]
What would you do next, after seeing/using this widget? [Written response]
Do you have any suggestions or other feedback? [Written response]
-
When asked what elements they think are clickable, testers said:
Logo/company name: goes to company website
Star ratings: goes to page explaining rating, review
Column labels: sorts rows
Price: goes to page explaining what full coverage includes
Review button: goes to customer reviews, unsure what kind of review to expect
Only dropdown is clickable
Which table do you prefer?
Pain points and other findings
All testers found the table useful, especially because rates and star ratings what they want to compare.
Preferred Prototype B over A. Most said having the company name next to the logo is distracting since the logos are recognizable.
What kind of review? Testers are unsure where the “Read review” button leads to. Will they see customer reviews or ValuePenguin’s review?
Want sorting. Testers want ability to sort the data by rating and cost.
Next steps
Update prototypes based off of tester feedback
Meet with engineering team to scope project and create Jira tickets
Collaborate with analysts to create datasets that will power the widget
Final Prototypes
Desktop
Mobile
Key Takeaways
Once developed, three different versions of the table were added to pages to test for engagement and SEO impact.
Bounce rate improved. Click tracking showed high engagement with the state dropdown and star ratings.
Happier team. With the new widget, editors saved an estimated hour of work per page.
Next steps
Time to scale!
Now the framework was built, the team could easily create similar widgets within insurance (home, health, etc.) and other products.