Project

Project Brief

Help users improve their driving behaviour by coaching the driver through their previous trips

Duration

8 months (January - September 2022)

Tools

Team

2 iOS developers, 3 android developers, 1 lead designer (me), 1 normal designer, 1 area product owner, 1 scrum master, 1 QA

What happened

We started the project with the business goals. Since our main goal was to acquire + retain clients we picked a solution which was marketable to clients. In the end this sales driven strategy did not help us increase or retain our clients.

What I would have done differently

Acquiring and retaining clients will only happen if we provide them with a valuable feature. Therefore we should have started by considering both the client's and users' goals.

Technical constraint

One example of many technical constraints we encountered during the project

Current design

This is a view the user sees when they are browsing one of their previous journeys. They can click on the pins to see details about their driving. The pins include information about harsh maneuvers such as braking, cornering etc.

Requirement

Show the user a pin for speeding as well as harsh maneuvers. This will help us give more specific coaching about speeding eg. “You were speeding in a school zone, beware children may run onto the streets suddenly.”

Constraint

Developers don’t know where to show the pin. What if the speeding was for only 1 km? eg. pulling into a motorway? Does autobahn count as Speeding? If there are more than 30 speeding pins should they be clustered into 1 pin? etc.

How I solved it

Got the right people together: created a task-force of a data engineer, a developer and a junior product owner
Created a rule eg. show pin if: 10% of the route contains continuous speed AND x AND y
Used existing trip data to simulate results using above rule + did a sense check + adapted where needed
Aligned stakeholders to new rule

UX compromise

This app is essentially trying to change driving behaviour. This is an incredibly difficult task. The accepted knowledge in behavioural psychology today is that behaviours follow a loop of trigger, action and reward (shown on the right).

There was a compromise made at the start of this project between Sales and UX Design. Sales had promised a “coaching feature” to our clients. Adding a coaching feature enables the user to take the right action in the hook model. However, the user will not take the right action if they are not motivated by the reward. Furthermore, the action won’t become a habit/behaviour if the user does not have a trigger.

Due to time constraints we could not work on all the aspects of the hook model. If we were to follow the principles of MVP or agile development then we should have created a closed habit loop first (no matter how basic) before continuing to improve upon just 1 part of the loop. In other words, we should have put our resources towards creating triggers + strengthening the user’s motivations (rewards).

This was a compromise I, as a UX designer, had to make in a sales driven product strategy.

UI Rationale

I am always growing and improving the framework/rationale which I base design decisions on. Here are some of the key concepts I used to make decisions in this specific project with a few examples:

Visual + usability principles: Hierarchy of information, cognitive overload, user freedom etc.
Insights from research: What did we learn from user testing, what does the user want from this screen? Guerrilla testing in the office, UX audits by colleagues, observations from SH?
Marketing: How can we market this product to our clients? Will the clients also be impressed? Can they get this from competitors?
App design principles: OS patternsQuantitative data: clicks, views, scores
Business goals/metrics: client satisfaction, safer driving

Outcome

❌ No new clients because of this feature
❌ The client to which this feature was promised ended up leaving

👀 Improve loss ratio -> long term goal which can not be calculated just by this feature
❌ No increase in telematic contracts + data
✅ 8% improvement in driver scores

🤷🏻‍♀️ Improvement in driver scores means more rewards for the user but whether this was valued is untested

My learnings

The company helps mid-size automotive insurers provide a telematic insurance policy (pay/earn rewards based on your driving) to their customers. This is achieved through Coloride; a white-label app which insurers can put their logo + colours on for their customers.
This strategy doesn’t work because the app is not customised to the unique needs of each client in their respective markets. This is proven by the fact that our most successful client (by far!) build their own custom app with us.
Legacy insurers don’t have a user centred or a digital mindset. None of our clients do any user research and had ineffective go-to-market strategies.

Lack of vision
Lack of transparency
Lack of measurements

It's extremely important to have a framework to base your design decisions on. This framework will evolve over time as I learn more and depending on the context of the project.
In order to advocate for the user I have to preemptively create UX allies in the organisation. They can support UX principles by backing up user centred arguments during key design reviews.

Coloride