Background
As part of the architectural requirements of Homeroom, teachers have to register in a school. In the first step of the registration process, teachers have to find their school in order to proceed to the next step where they enter their personal information. Therefore, the school search function and its user experience are very important for successful registration.
As a backup, If teachers can not find their school, they click on a link to skip the first step and are directed to the page where they enter their personal information. The link for the alternative path only appears when the system throws a ‘no school found’ error. When teachers click this link and complete the registration, EVERFI registers them into a generic school (under the hood). It is important that EVERFI does not lose teachers who cannot find their schools. However, EVERFI needs to contact and connect them to their school afterward.
What was the problem?
1. During the discovery phase, I found out through Google Analytics that 35% of teachers’ registration sessions threw a ‘no school found’ error. But, of the 35%, only ~10% clicked on the link to begin registering into a generic school. This meant that we were losing teachers who enter the registration process, cannot find their schools, but don't register to a generic school either.
2. During the usability testing with new teachers, I observed that none of the teachers who could not find their schools did not end up clicking on the link for generic registration.
3. Furthermore, the usability review of the registration experience revealed an interaction issue; since the link to generic registration appeared when the system threw a ‘no school found’ error, depending on what teachers typed the link could keep flashing which made the experience very unpleasant. See the issue below:
Why Did It Matter?
The goal of this project was to find out what hurdles were in teachers' way, and remove them while improving the interaction to maximize the conversion rate. Registering into a generic school is very important for those who cannot find their school or don’t have a school to register into, e.g, homeschool teachers and schools that don't exist in our database.
What Was the Approach?
I created two tests to understand and measure which solution yields a higher conversion rate and a better user experience.
Test 1 - Omnipresent Link
For the first test, I made the link to generic registration omnipresent. The premise was for teachers to be able to find the link easily if they could not find their school using the school search. The drawback to this design was the possibility of teachers clicking on this link before trying to find their schools.
Test 2 - Generic Link in the Search Field
For this test, I designed the link to appear in the Search field when users start typing; teachers will see an omnipresent link as they type to find their schools. This pattern also allows collecting teachers' school names before they go to the second step of the registration.
What Did We Find?
Test 1 - Omnipresent Link: The number of teachers who clicked the link increased and that resulted in a higher number of registrants. This change resulted in an increase in the number of teachers in the generic school with an operational dept requiring users to be manually placed into their actual schools. The impact of this change was a 4% increase in the conversion rate.
Test 2 - Generic Link in the Search Field: The number of teachers who clicked the link increased slightly followed by an increase in conversion rate of ~2%. Since we were able to collect teachers' school names, the operational effort to place users into their schools decreased.
What Is Next?
The conclusion was to design a solution that helps us achieve a higher conversion rate while keeping operational costs reasonably low. This called for borrowing ideas from both designs to achieve our goal.
Please check back to see the result of this test!