What is it?

A branded e-commerce site for selling WiFi data packages to drivers across the EU.

Who's it for?

Drivers for whom fast reliable internet on the road is vital.

Automotive Webshop

Improving a product I didn't really like

This product was launched before I joined the team in 2016, and a proper design process was not followed creating an interesting situation: How do you make a success of a product that is based on what I always thought was a fundamentally flawed idea: Why would I want WiFi in my car when I have my phone? Seriously, when I started on this project no one could answer that question for me. Whenever I talked to friends or family about the product they asked me the same question, it felt like it had been conceived by someone locked away Monte Cristo style and prevented from discussing the idea with a single human being.
‍
First things first: Research. I set up Google Analytics and Hotjar to help analyse our users behaviour. We also had a PowerBi dashboard that was underused. I had the data team put together a custom view for the UX team to showcase the key statistics that we were interested in. There was no real concept of KPI’s beyond data plan sales at that time, but I wanted to focus on user growth, overall data use and data use per user, all of which would lead to better sales in the future.

The first issues to tackle were the most obvious. The homepage for this consumer facing data service was… a login page. There was nothing that explained the benefits, nothing that guided users through the complex setup process, no information on pricing. You don’t need to do reams of user research to understand why growth was stagnant. I did however talk to any driver I could find about the appeal of the product. They were all nonplussed. They asked why they would need WiFi in their car when they have a mobile on them at all times. They didn’t see the benefits of the product as particularly beneficial. The signal is stronger, but that’s not a major concern on the road. You can stream content to 8 devices at once at a high speed but the high cost of data makes that unattractive. You can use any network in Europe, which makes it somewhat useful for someone that drives across borders regularly, but if you don’t fall into that (tiny) category what is the point?
‍
The biggest shame of all was that while the concept was hard to sell to people the underlying product was very good. If you switched from mobile internet to the in-car WiFi it was significantly faster, the signal rarely dropped and the network was strong and stable enough to stream HD video to as many devices as you could reasonably want. I also discovered that while the product was pretty pricey in an Irish context it was cheaper than mobile data in Germany (our largest market).

I started the design with pen and paper developing through wireframes into a fully fleshed out inVision prototype that I could review with drivers. As I worked on this I also held meetings with our procurement team, putting pressure on them to renegotiate our data contracts and get better prices (We got a 20% reduction, which made a big difference in our price competitiveness).

The original proposal designed before my involvement.

Development

On the UI design I worked in an Agile environment with our dev team with design running one sprint ahead of dev to ensure that blockers were minimised. We organised our work in Jira and used Invision for handing over assets to dev. The challenge here was to build four separated branded versions of the site at the same time (VW/Audi/Skoda/Seat) finding a base design language that was consistent enough to make development and the future onboarding of other partners easy while giving the partners enough flexibility to feel like their branding was fully on show. 

The new version of the site launched in January of 2019 and we saw a change from a small stagnant userbase to steady, if unspectacular, growth that gradually increased over time to the point where now sales have increased by 300% compared to this time last year, and that’s in the context of a global pandemic where people are driving far less than they were before.
‍
This growth is not entirely driven by the 2019 redesign. Our last major release in 2020 brought in a free trial feature for Audi and VW users. We identified this as a priority because of the aforementioned disconnection between user expectations and the quality of the product that showed in my user research. It took a while to bring to fruition (No one wanted to pay for it), but when we were finally able to get it designed and developed in late 2020 we saw an additional boost in user & sales growth.

Download Prototype

End Results

Monthly Sales Growth of 750% (600 p/m to 4,500 p/m)
Active User growth of 430% (3.3k to 14.2k) 
Data usage growth of 400% (4.1 TB to 16.6 TB)

Extra Bonus Failure Anecdote!

I had a month or two when the product wasn’t being given much dev resources and with no pressing work to do I decided to investigate the registration flow. Our on-site registration flow was quite standard, account creation, link your vehicle and complete an ID check (A legal requirement in some of our markets), but when you take into account the process of setting up and connecting to WiFi within the vehicle it became a very difficult task for users and a very difficult process document as there were so many different vehicles to cover.

I conducted a number of straightforward user tests, I sat with them in a car and asked them to set up the WiFi and create an account. On average it took users 45 minutes (!) from turning the car on to connecting to the WiFi. Trying to solve this problem lead to so many of dead ends. I tried to work with our partners on the parts of the flow in their HMI (the in-car computer display), redesigning it to be simpler and more efficient, but found that in most cases HMI’s can’t be updated after the car leaves the assembly line. With that option out, I instead produced a series of videos for each vehicle guiding users through the registration process.

In user tests I found that after watching a 2 minute video users could now complete registration in under 10 minutes compared to 45 minutes before. I produced a few of these (there are so many vehicle variants it’s hard to produce them all without a proper budget) and shared them with partners, hoping that they would link to them through their portals and improve registration that way. They didn’t. It may seem odd to include a big dose of failure in a portfolio (please employ me), but I think this communicates as much about my approach as my (many) successes.