From one of the rounds of regular user research we've got the following recurring problem:
The existing placement test was based primarily on grammar knowledge, required significant time investment, and had some UX issues.
Confidence level survey
Together with instructional designs from didactics, we conducted a survey to see if learner's confidence level can realistically reflect their level of language knowledge.
Wireframes & user feedback
A/B test results
The Can-do placement was launched as an AB test as a part of the registration&sales before the sign-up.
As a result, we saw a small decrease of 3.5% in the Visitor to Registration conversion, but a big increase of 22% in Visitor to Sale (Sale = subscription purchase) conversion.
Based on the AB test results, we're moving the placement post-registration and testing a version with its integration in the Learning path.
© 2022 Natalia Volgina