At-home return material rate (RetMat) was only 40%. Meaning customers who bought $95 at-home kits were not returning their impressions.
Candid had two core pipe lines of business at the time I worked on this project: the at-home, do-it-yourself option and the studio location option. Our studio locations were doing pretty well since customers could get in and out in ~30 min. They'd show up and get a 3D scan of their teeth and some photos taken so an orthodontist could create a treatment plan.
The at-home side was a little more complicated. Customers had to order a starter kit to get started with treatment. They would have to complete 8 intraoral photos of their face and mouth and do 5 impressions of their teeth. They would then have to send those in for our lab to scan, check for quality, and then pass on to our orthodontic team to use in creating customized treatment plans.
- Me (product design)
- Sudy, head of research
- Mary, product manager
- Joseph, front-end engineer
Customers feel like doing impressions is daunting and the printed isntruction booklet don't give them the confidence to do so. By creating digital instructions we can provide more nuanced detail around the process. We believed by doing so we sould 1) improve our RetMat, 2) improve first time photoset and impression approvals, and 3) iterate quickly on the best way to present instructions. (Plus paper is hella expensive if you're changing your instructions every 2-3 months)
Our approach was one that isn't always normal in an early stage startup. We wanted to make this experience right from the start. So we decided that we would take the time necessary to crack the problem by running moderated, in-person user testing. In order to at least stay on par with reality, we had one of our lab assistance on-call to review photos and impressions during the process. We would rate our user testers based on the acceptance criteria of the lab.
At first, we thought our customers needed a little bit more hand holding and decided to take a page out of Aptiv's playbook. Let's create an app experience where they enter the flow and it guides them through each step and provides a timer when they need to do a time sensitive piece of the process. We took this prototype and ran our user tests with 6 people. Unfortunately, it didn't work out so hot. Most of our testers failed to meet the standards of our lab. So we went back to the drawing board.
We noticed (and got a lot of feedback in post-test interviews) that the users felt rushed. We altered our approach to focus solely on using video. We rescripted and reshot our instructions to be a 12 min video they could scrub through, pause, and rewatch if needed. With the next batch of testers, we saw improvement over our last round of user testing, but still didn't meet the criteria. Based on the responses during post-test interviews, users felt a little overwhelmed with how much was crammed into the video. So we altered our course once more.
This time, we split each piece of the video up into their own sections. Users would start with the photo video, be walked through photo taking and uploading, then move onto doing just their top teeth, and once done with the 3 top teeth impression trays, we would move them onto doing their bottom teeth (a total of 2 trays). Each video also had examples of poorly done photos/impressions so users would get more detail into what to look for when completing the process. With this last user testing group, we met our lab acceptance rate and standards.
As always, the job isn't done until we've tested it en mass. We ran a 50/50 A/B test for 2 months in order to get to a statistically significant result for our conversion rate (RetMat). Meanwhile, a few of the engineers not working on the project were skeptical and decided to place a bet with us on how it would work out. A bucket of chicken nuggets was on the line for each person placing a bet. This was go time!
My team was expecting that we'd at least net out neutral, meaning we could ship it and at least use it as an excuse to rapidly iterate. Instead, we won a bucket of nugs for each of us.
Our RetMat rate went from 40% of all starter kits purchased to 60% of all starter kits purchased!
We improved RetMat by 20pp!