We’re Electric
Product Design Case Study
User Research
An Unexpected Survey
Even though I felt fairly convinced a car-hire service offering exclusively electric vehicles would be a valuable service, I resolved to make sure by doing some unplanned quantitative primary research before conducting any user interviews.
I approached members of the local tennis club via WhatsApp, as subjects for the survey.
Testing the EV use-case by survey
The secondary research had given me increased confidence in the hypothesis that an all-electric service would have market value. Many dominant car hire services were now providing a filtering option specific to finding electric vehicles - but there were limited options, often resulting in ‘0 results’ for specific locations.
Other speculations during orientation had truth to them: some of these services were themselves adapting to create more flexibility through experimenting with car-club style business models (SIXT, Enterprise) which allow drop-off in different locations than pick-up but only within defined urban boundaries.
However, making the client’s brief for a car hire service into an electric car hire service still represented a significant diversion from the initial assignment and would mean I would need to significantly adapt the brief set by the client (which in my case might affect my grade in the diploma!).
Therefore, even though I felt fairly convinced a car-hire service offering exclusively electric vehicles would be a valuable service, I resolved to make sure by doing some unplanned quantitative primary research before conducting any user interviews.
Since this was a low-budget project, the opportunity was not available to validate the interview findings themselves with a survey that had a large enough volume of responses for a statistically significant validatory result.
Whereas there would be value in getting just a small amount of survey responses to get a signal that might confirm that narrowing of the product in the envisaged way was viable.
Cost-benefit (unusual order for doing a survey)
Overall, conducting this survey before the user interviews survey felt like it would be a small amount of extra work for a large amount of research value:
A survey would provide real data that I could triangulate with the competitive benchmarking and the subsequent interviews to galvanise confidence and gain clearer direction in the proposed all-EV use case. (That, or it would invalidate the all-EV hypothesis early in the process and before the interviews, which would be economical as a process of elimination).
Also, being without access to budget to generate a volume of respondents high enough to bring scientific significance in the survey data, maximising the number of reliable data sources felt important for convincing potentially resistant stakeholders that my hypothesis was validated enough to justify building a service focused on electric vehicles exclusively.
The survey was also useful to conduct before the depth interviews because choice of candidates for interviews and the composition of other materials later in the process, could be usefully influenced by the slightly different target market for a green car hire product.
This survey would validate (or avert) the all-EV approach, giving clarity for how I should approach those subsequent choices.
It didn’t seem from the results like there was a heavy weighting either way here. With such a small sample size, I didn’t let this consideration influence the final design in the end.
Knowing that my sample size was small, I adapted the conventional Lickert scale to be 6 (rather than 5 or 7) in order to force users to dwell on whether they felt the positive or negative side of neutral about this question.
Approach to survey composition
This survey became long and complex and I wanted to use it also as a method for sourcing a person to do a usability test with. This became a dependency that began delaying my work.
Since I have many contacts but am unsure who does and doesn’t hire cars, I decided not waste the opportunity to use the survey to also find out who has hired at the same as gathering information about attitudes to electric cars.
For example, I have used logic flows in the survey tool to harvest attitudes toward hiring cars and toward the environment from anyone who has not hired a car. I have used logic also to ensure it remains a short survey regardless of the logic path that each user takes based on their answers.
In the ‘has hired’ path, I added asking if they mainly used the car for their journey to their destination because this is a really important aspect I picked up from the benchmarking and the note taking, and if there was a strong signal here, it would likely influence the design of a future service. For example, in the era of competition from easily accessible ride- sharing apps, an affirmative signal from the data of 'mainly used the car for their journey at their destination' would mean a UX that works somewhat like a ride-sharing app but allows you to drive yourself.
Such a UX would however be less relevant in the case a car was used as the primary means of transport to the destination. The logic paths this question creates should allow digging into each use case to glean useful insights for an electric service.
So, I kept the question about use of car in, as it could lead to insights related to behaviour of use, to inform design decisions for an electric car service.
I chose Typeform as the survey tool as it has a friendly and conversational UX.
I was also able to use the free tier because I wasn't seeking more responses than that tier's limit (10 responses).
Link to the finalised survey: Car Hire Habits
There were 5 responses (excluding my test response). Although this sample size is very small, the unanimously positive response to the survey’s main question was a strong signal in favour of my envisaged use-case of a car-hire service that only offered electric vehicles: everyone thought to some extent about the climate when making decisions about transport. I took this response, together with the answers relating to barriers to hiring EVs, as an qualified ‘yes’ to my hypothesis:
There is a gap in the market for a car-hire service that provides electric vehicles exclusively and that educates the user about how to overcome the common barriers to EV uptake: battery range anxiety and vehicle charging process
User Interviews
I sourced two interviewees from those I surveyed in the previous step, in addition to the two supplied by the client. I invited these new interviewees by email and conducted one interview in person and one remotely. Each included a comparative usability test.
According to an approach of going broad into the potential competitive landscape rather than getting overlapping feedback about the same services, I interviewed my interviewees on two of the services I had looked at during benchmarking (Europcar and Turo) rather than going over the same services as the interviews I received from the client (SIXT and Green Motion).
I thought it would be interesting to see the response to Turo’s relatively new P2P business model.
I also wanted to gather more data on major services (Europcar) as Green Motion hadn’t been impressive during the competitive benchmarking and seeing another international service akin to SIXT would be worthwhile, as these kinds of service would have had the most influence on users’ perception of a what is ‘typical’ for a car hire service.
Picture: attachment to email sent to interviewees, explaining advance setup requirements
Picture: highlights I generated from the two already-recorded video interviews and comparative usability tests supplied to me by the client (covering SIXT and Greenmotion)
Picture: screenshot of my interview with Darri (Turo and Europcar)
Approach
I organised a comparative Usability Test and Interview (for mobile, in person), with a friend who had recently hired a car for a holiday in Spain.
Objective: investigate UX for hiring electric cars specifically, as well as overall usability. Find out how UX of newer peer-to-peer site Turo (uncovered during competitive benchmarking), compares to an existing mainstream offering.
I chose a ‘typical’ car hire company (Europcar) and also a newer peer-to-peer car sharing site (Turo), based on my hypotheses from the benchmarking. I made one main deviation to the client's prior interview guide in that instead of asking the interviewee to find an automatic vehicle, I asked him to select an electric vehicle.
Interview with Oscar (time problems!)
Unfortunately the interviewee needed to push the start time back owing to a work commitment, so I skipped the background interview and moved straight to the comparative usability test. I then rescheduled with him and completed the background interview remotely over Zoom with LetsView as the mobile screen-sharing app.
In the comparative usability test: I was using Zoom with Firefox and the thumbnail views worked differently (they were floating windows instead of fixed inside the Zoom window).
This meant I couldn’t slide them in to increase the size of the interviewee’s video, I needed instead to increase the size of the floating window and I since noticed there is a small overlap in the video download of the call (this wasn’t overlapping in the live call). This obscures an important area of the screen (top right), including one thing the interviewee directly comments on.
So for the record: the filter icon top right corner of screen is replaced by a number when at least 1 filter is applied. This means Oscar was slightly unsure where to find the icon again when wanting to apply another filter.
I captured this in a sticky.
Interview with Darri
This felt like a good interview. Not only did I manage the technical aspects better, I also led with a better introduction as there were no pressing time constraints this time.
As a result, I felt I got better insights from interviewee Darri, than I did from the Oscar.