Why did Go-Jek turn to us for council?
In May 2016, Go-Jek, Uber’s biggest rival in Indonesia joined the big league when it was valued at $3 billion. Recent years had seen a phenomenal growth in Go-Jek’s customer base as well as their team. Their driver fleet had spiked from just 2,000 drivers to over 250,000 drivers in little over a year. This huge fleet of drivers was responsible for rendering a bunch of services ranging from motorbike taxies, food, grocery and courier deliveries to bringing beauticians and masseuses to the users’ doorsteps, all through a single driver application. To keep up with the growth in both their size as well as service offerings, Go-Jek needed a reliable yet scalable app for it’s drivers.
The driver app used by our 250k+ drivers had not received a refresher since it was initially launched. This meant that every time we wanted to introduce a new feature or add information on a screen, the experience would worsen. We ultimately needed a scalable design for this ever-changing app.
Product Manager, Go-Jek
Why did we take on the challenge?
The app had to be dead simple, because a lot of drivers who are also first time smartphone users depend on it for their daily bread. However, the fact that this one app was required to render 6 different services made achieving this simplicity extremely complex. To understand how we could simplify this complex app, we worked very closely with the Go-Jek product and development teams. After deep diving into Go-Jek’s challenges for a few weeks, we knew our designs had to counter some tough business problems. Read on to see how we solved them.
How might we create a fast experience for Go-Jek customers without forcing Go-Jek drivers into accepting orders haphazardly?
Go-Jek’s data suggested that 10 seconds was the optimum time duration for a driver to accept/reject a bid without making the customer wait too long before her order got confirmed. But 10 seconds was still too short a timespan to comprehend all the information about an order and make a well informed decision.
We sat down with Uncommon to review what were the key pieces of information to have when presenting an order to a driver. Drivers care about how much they will earn in fare and in bonus. And they care about the location surrounding this order — for example, whether the destination is near their home because they are just wrapping up for the day and want to head back.
Product Manager, Go-Jek
To get a better understanding of what was going on, we evaluated the time required to comprehend (and act on) each piece of information on the bid screen.
We resolved the problem by stripping down the information to a bare minimum, without eliminating information critical to decision-making. For example, instead of showing the complete address of the pickup and drop points to the driver, we decided to just show their broad locations.
How might we dramatically reduce cancellations once an order is accepted by the driver? (CTA)
Because a lot of Go-Jek drivers were first-time smartphone users, accidental presses were common, and positive call-to-actions (CTAs) were often confused with negative ones. The additional pressure of using the app on the go in direct sunlight and street noise further added to the confusion.
For those reasons, we chose the time-honoured slider interaction for positive CTAs, which was introduced on almost every screen of the app. Right after first use, the drivers associated it with an interaction that would lead them to the next step/screen, without paying any attention to the CTA’s colour or copy. Because the interaction demanded a deliberate action, accidental presses took a sharp dip, resolving one of Go-Jek’s biggest concerns in their driver app.
Try out the prototype yourself –
How might we reduce the order completion time by reducing the time spent by the drivers on their phones?
We helped Go-Jek reduce the average order fulfilment time by reducing the time spent by the drivers on their phones during the journey. This was made possible by –
1. Introducing faster switching between overall journey view and current leg view
In case of multiple order deliveries, the drivers needed to take a peek at their overall trip without exiting their current trip leg. Here we implemented strong motion design which allowed the drivers to stay on their current leg but also see the whole journey by simply swiping up the trip card.
2. Removing blocking ux patterns throughout the app
The original app was designed to block the driver in some cases (to avoid misuse/confusion). For example, the drivers were forced to upload a bill in case of a grocery order: until the bill was uploaded, the driver could not proceed further in the app (because a difference in the actual bill and the estimated amount could affect the amount to be paid by the customer). This was an obvious problem because the driver could remain stuck indefinitely due to poor internet connection.
The new design solved this problem by uploading the receipts in the background, allowing the driver to proceed to the next step and see the delivery address. By the time the driver would reach the customer, the bill would be uploaded (and the customer updated with the amount to be paid).
Results: The final outcome
While most startups understand the value of design in acquiring and retaining customers, very few startups understand how design can play a role in delighting and retaining their own employees. Go-Jek definitely belongs to the latter group.
For Go-Jek drivers, being a driver on the platform is their living. They are able to earn a steady income, provide for their family, and have the flexibility in working hours. Throughout the redesign, one thought we constantly had was that we need to start treating our drivers like we do our customers and striving to provide a great and delightful experience for them.
Product Manager, Go-Jek
Besides making the lives of the drivers simpler and more delightful, the new app has successfully improved the following business goals –
- Faster order acceptance rates
- Faster order completion rates
- Reduced order cancellation rates (by the drivers)