There is a misconception among drivers that working during a surge is the best-earning strategy, especially in some mature markets. The perception was validated within the 2-way forums* and reflected with the actual driving behavior we observed in quantitative research via product analytics.
We also observed a segment of drivers carrying a physical notebook and manually logging the fare each time they drop off their passengers.
From these observations and behaviors, we realize we have an opportunity to empower drivers to strategize better to earn more.
Introducing Earning Insight Widget
For the segment of drivers with manual logging behavior, the idea is to have a reliable and robust widget that displays earnings per hour that is tally and consistent with the fare recorded in their logbook. Another crucial "Jobs-to-be-done" here is to showcase and "debunk" the myth of driving during a surge is the best-earning strategy.
One of the hypotheses we have was giving insight to drivers and being transparent could be the vital first step of gaining trust to nudge for behavior change.
Then at the same time, nudging for a new behavior shows how others are earning with a different strategy like driving outside the rush hours during a surge.
Concepts & Explorations
For this project, we did Usability Testing with testers who are real drivers inside their cars. There are key learnings & takeaways that we later implemented into our final design:
- We break down the screen into three pages with 8 hours each. This design treatment matches most use cases regarding having shifts instead of showing a 24 bar graph on one screen.
- Showing all 24 bars can be overwhelming and also harder to tap on the bar itself.
- In the exploration design, the left and right chevron icons are "moving" as the text changes. For example, when cycling through the shift hours from "1AM - 8AM" to "9AM - 4PM", the left chevron seems to move slightly to the left and right.
So with all the key insights and learnings, we implemented a couple of changes in the design we finalized for MVP.
With all things considered, if our hypothesis is true and potentially helps drivers to earn better, the widget might also nudge them to change their current routine, like waking up earlier or drive longer hours slightly.
How do we balance nudging for earning better while not pressuring drivers to drive longer, resulting in fatigue and ultimately having an unintended, negative impact on their safety?
Behavioral Science to the rescue
One major thing that makes me feel privileged and lucky to be on the Economics team back then is that I work very closely with our only Behavioral Scientist, Preeti Kotamarthi. Together, we initiated and introduced the Nudging Responsibly Guiding Framework as a bottom-up initiative. We intentionally didn't call it an "ethical" framework, given there was no official authority, and claiming such responsibility can be counterproductive. It could invite unnecessary and unwanted scrutiny, which could distract and derail us from the real issues at hand.
Nudging Responsibly Guiding Framework
The framework consists of an audit with guiding principles and questions. From the audit, teams can readjust and make the necessary changes to be more responsible with the nudges introduced in the design. So how do we apply this framework in the context of Earning Insight widget? Let's have a look.
Does the beneficiary have a choice?
We added an on/off toggle as choice architecture.
Now, the driver has a choice to turn it off if they wished not to see it.
Does the beneficiary feel pressured?
We changed the widget switch to be turned off by default.
Even though it might seem we are risking lower growth in terms of success metrics, we believe putting that choice on drivers is the right thing to do.
Some of the high-level successes of the project are:
- Observed an increase in the earnings per hour for drivers who utilized the widget.
- Positive qualitative feedback, such as asking to keep the feature permanent.
- We received feedback asking for more insights and information such as show earnings after commission, pickup area comparison, and more.