Personas
Our SEO consultant team identified three key audiences through keyword research. We began by using analytics data to verify our personas.
This is what we found:
- Car Buyer – research (decision support)
- Enthusiast – news announcements, fresh reviews
- Lifestyle Consumer – light-hearted and/or viral, consumable content
Through my own qualitative research by conducting user interviews, I started to realise that automotive content had a large breadth of interest. Ranging from a fairly broad “I enjoy reading about cars” to incredibly granular interests ranging from “classics” to “racing cars” to “hyper cars”.
Performance Benchmark
Part of my major responsibilities was to carry out a performance benchmark assessment. I developed a two-part test usability test conducted on-site with 8 users. What I wanted to focus testing on were:
- Areas of the current website that were performing well
- The ease of use against a direct competitor’s website
Choosing to involve the project team in the usability testing was important in creating an environment of shared understanding.
Conducting usability testing early gave us critical feedback on what elements were working to support their goals, and also identified opportunities to capitalise on gaps in our competitor’s user experience. We also had participant’s responses supporting the theory that good visual design created the illusion of “easy to use” and favoured by our participants.
information architecture
Armed with the knowledge that competitors had pages set up for indexability, we had to work fast to prevent the risk of falling too far behind in Google search rankings. The SEO consultant team reworked the information architecture to meet the following objectives:
- Landing pages with SEO-friendly URLs
- New content category “groupings” to target the Car Buyer, i.e. Sedan cars, Eco-friendly cars, etc.
I worked on creating a Treejack test through Optimal Workshop. My scenarios reflected our research on user objectives, such as finding car review videos. Our results came back with a 74% pass rate, with 64% directness, from 57 participants.
A deeper analysis into the results confirmed that our architecture was structurally sound. Our lower than expected success rate was attributed to two scenarios where we hadn’t anticipated users using a different path to achieve the same goal. Another constant reminder that we were not mind-readers when it comes to predicting usability!
Wireframes
Thinking back to our competitive analysis, it was time to pick the best thing our website would do. The close association with a household automotive sales brand made for a strong argument to focus on research material for car buyers. The pull of enhancing conversions and monetisation through use of strong review-style content was too great, and the decision was made.
Our client wanted the website to be deployed in record time… Our research was great but they found a cookie-cut WordPress theme template they insisted on using. This was a constraint I would now work within.
A big part of the design challenge was in distilling our findings to a solution in a meaningful way. Just looking good wasn’t going to make any waves!
Using UXPin as the platform, I worked towards the following objectives:
- Incorporating functionality to support content discovery for the Car Buyer market – i.e. “make and model” search
- Consider the current behaviours of existing users – not force them to relearn all their browsing habits – i.e. keyword search
- Prioritising content preferences using persona and research insights New “All-makes-All-models” landing environments (content hubs)
- Generate page-views – as many as I could without breaking the experience (no Dark Patterns used here!)
Generating Page-views
The website was created to generate income from advertising, equalling page-views as a success metric. Not harming the experience wasn’t enough, it had to impart some benefit to our users. We found a plug-in that allowed us to create content “tabs” to act as a tertiary navigation system – this allowed users to tab to relevant content such as images or videos. As they moved tabs, it would generate a unique URL.
This solution met our objective for page-views and produced benefits. Let’s say you wanted to evaluate a car based on its’ aesthetics. You could now get to all photos of the particular model through a simple click of a tab. Because the whole page did not refresh, it felt like you never left the page.
The unique SEO-friendly URL generated by each tab made the content easily indexable.
You could now get to all photos through a simple click of a tab. Because the whole page did not refresh, it felt like you never left the page.
“How can we increase advertising on-page?”
During our research, users complained of the heavy-handed advertising on the previous website distracting them from their tasks. It was an issue.
Our participants understood that advertising was part and parcel, and some even commented that if relevant, they found ads useful. I wanted to move from “standard spots” for advertising to ad spaces becoming seamless with editorial content.
Playing around with logic such as display ads for the newest family VW on the Volkswagen or Family Car page would increase relevancy with the probability of higher click-through rates.
Usability testing
I conducted more usability tests during the wireframing process. As much as possible, I was incorporating multiple ways to discover content.
We tested 6 participants with our prototype and saw an improvement of 16% in task success. Our user’s self-rated score of “ease of use” jumped from an average of 6.4 to 8 out of 10 (very easy). Our overall performance was looking healthy. Even distraction levels of advertising stayed at a low average of 3 out of 10 (very annoying).
