RepairPal

Abstract

An initiative aimed at reducing consumer cancellations for auto repair appointments booked through RepairPal.com, this project spanned the length of 2 quarters, with ample time for research, discovery, data collection, design, and implementation. Other adjacent goals included learning why consumers and shops initiated cancellations while improving data fidelity around appointment cancellations (an area that was formerly a large, black hole with questionable data).

Role

• Sr. Product Designer - Came up with hypothesis & A/B/C test concepts, designed solutions, and iterated post launch

Initial Data

At the beginning of this project we didn't know enough about cancellations to come up with any accurate solutions. Instead, we started with our own data to form a set of open questions we wanted answered during the discovery process. The prior year's figures told us the following: 40% of appointments booked resulted in an invoice, but only 4.4% were actually canceled. This indicates there was a discrepancy between appointment attendance & completion, and actual billing.

Open Questions

• Why were consumers booking appointments and then not showing up? (An inference)
• Why were only 4.4% of appointments being canceled if 60% of appointments booked DID NOT result in an invoice? Why were consumers/shops not canceling?
• How could we learn more about why consumers were canceling and work to ideate solutions to reduce overall appointment cancelations?

Areas for Opportunity

• Understanding what drives consumer initiated appointment cancelations
• Understanding what factors lead a shop to cancel an appointment
• Improved data and understanding around what we know about our existing customers (shop & consumer)

‍

Primary Research

From our open questions, we then sought to understand: What user motivations drove appointment cancelation across the RP ecosystem (whether the cancelation is shop or consumer initiatied)? Can we isolate which cancelation reasons have the highest rates of recurrence amongst our audience in order to retarget to those consumers, prevent them from canceling, or offer a different shop to book with?

For primary research, two workshops were run: the first being an initial analysis of all existing consumer cancel reasons (especially if 'Other' was chosen or the user wrote in their own reasons for cancelation) with special note given to new categories we needed to create, and the second being a workshop comparing the different motivations for canceling appointment between consumers and shops to see if we could mend this bridge from an organizational perspective (ex. a member of the shop team reaches out to a shop if they keep canceling their appointment for labor or parts issues). In order to solve for the most recurrent cancelation reasons, we first needed to know what they were.

Workshop 1

Workshop 2

‍

Next Steps...

Once we understood the data more, we were able to understand where we needed to start. We also created a list of new appointment cancelation reasons to incorporate in the design to improve the data from cancelations over time. We were also able to prioritize which workflows we wanted to start testing with on the consumer & shop side.

‍

When it came to testing & design, we wanted to take an iterative approach, reducing developmental overhead, especially if we could make smart learnings earlier rather than later. We tested multiple approaches centered around reframing the user's primary action via hard nudge and soft nudge approaches. The hard nudge is a more overt approach whereas the soft nudge relies on small tweaks made to the copy & UI to reframe the consumer mindset around canceling while they're actively trying to cancel their appointment.

• Goal: Redirect user behavior from cancelation to rebooking and understand why consumers are canceling so we may ultimately reduce overall cancellations
• Hypothesis: By prioritizing "Rebook" as a primary action, we can increase rebooking conversion. We can also promote higher fidelity of data collection around cancellations if we provide more specific options for users to select.

‍

What we learned...

The Good

Below we studied the conversion of users who went to cancel & ultimately rescheduled their appointments. We saw the following:

• 26.24% of consumers who chose "I missed my appointment" as their cancellation reason clicked the link to reschedule
• "I missed my appointment" was the reason for 1.40% of consumer cancels in April & May, which decreased to 0.77% of consumers cancels in June
• Due to the success of initial testing of variant 2, all cancel screens were moved to soft nudge version 2
• Tested "hard nudge" and saw an increase in cancelations in July and August back to 1.43% and 1.23%, returned to "soft nudge" in September & removed ability to cancel past appointments, "missed appointment" cancels went down to 0.42%

The Neutral

• Free text responses remained relatively the same at about 30% through July.
• Modified wording and categories in late July and saw use of other decrease in August. Free text responses went down from 30% to 15%

‍

Next Steps

• Move the financing flow to a soft nudge (i.e. the consumer is canceling for financial reasons)
• Add retargeting emails for 'emergency/scheduling conflict', 'needs a dealer', and 'shop can't perform this service to recapture this lost appointment traffic and reduce waste
• Add a free text response for 'problem with shop' to gather more information for the shop team
• Reorganized cancelation reasons to prioritize the most commonly used and update category/subcategory alignment again on 10/2
• Removed "other" as a consumer cancelation option on 9/18.  Overall cancel numbers didn't change, but we saw increase in usage of "the shop was too far away" and "emergency/scheduling conflict" (both almost doubled)