Save and compare
Overview
The customer need for a car comparison tool has continually surfaced during various Auto Trader research sessions. Many customers currently find ‘hacks’ to fulfil this need e.g. opening multiple tabs displaying advert pages in their desktop browser, and printing out adverts. There had a been numerous attempts at compare through the years at Auto Trader, all removed from site due tp low usage. But time and time again the theme of comparing cropped up in our interactions with users. Around the same time a commissioned piece of consumer research also surfaced the theme of compare, with a slight twist. The theory was called an ‘anchor car’, this was a car currently owned by a user and being used to compare against possible new cars. For example, ‘I need a car with bigger boot and this one has a much bigger boot compared to my current car’. Armed with this research we set out to investigate compare for Auto Trader, using the anchor car theory as the starting point.
Problem statement
At Auto Trader we have a wide range of filters to help users refine a search down from 500,000 cars to a handful of viable options. But often users struggled in whittling down that handful of options, often hacking their way to compare but printing out adverts, using multiple tabs and sometimes even creating their own comparison spreadsheet.
Users and audience
Car buyers on Auto Trader
Roles and responsibilities
Lead product designer for the project. Working alongside a product manager, tech lead, a team of developers, front end developer and QA. This project included close collaboration with the research function within Auto Trader.
Scope and constraints
There was an initial recommendation that this was a feature solely for registered users of Auto Trader
Our process
Like all great theories the ‘anchor car’ theory came to a crashing halt almost immediately into the process. Working with the research team, we set a number of users generative design and mental mode tasks. It soon became clear that the ‘anchor car’ had already been discounted by the time a user had got to the point of whittling down a list. Search filters, reviews and general make/model research had done the job of demonstrating what these options did better than the users current car.
Compare journey map
What also became very clear through the generative design task was a clear theory of a ‘basket’ or a ‘list’. Users already had a mental model based on their e-commerce experiences, creating lists or baskets and from those lists making a decision on how to proceed. This felt like the best place for Auto Trader to help these car buyers. Users already could save cars to a list on Auto Trader but the feature was behind registration and was hidden in the depths of the users account area. Despite this, saved adverts was the second most used navigation item on our desktop experience, not bad considering the only 8% of users on site are registered. As we investigated more into users who saved, there were some startling statistics.
Generative design user sessions
A user who saves is a much more engaged user
They viewed more adverts per session: 8.53 vs 3.53
They viewed more adverts per user: 12.08 vs 7.30
Their average session duration was much higher: 18:50 vs 07:43
Plus the number of cars being saved in a week on Auto Trader was an eye opener considering those who were able to
Mobile Web: 61k
iOS: 402k
Android: 239k
Saved advert lists
This was bit of lightbulb moment for us, how could we bring this useful feature to more people and weave compare into it, to fit with the mental model our users had described.
Over the next month, a number or ideas were designed, prototyped, iterated and tested with users. Looking at how users initiated compare and also the tool itself, the learnings we took were invaluable. Themes emerged that married up with our initial findings and we were able to dig more into the ‘what’ users wished to compare on, looking at personalised views of compare and also a ‘versus’ mode, which presented in clear accessible language which car was either bigger, faster, more efficient etc…
Through these studies we got a place where we felt we had a solid first phase for the project;
Would be launched on mobile web initially
Saved adverts (and searches) feature would be given their own top level navigation item under a ‘Heart’ icon
Save call to action renamed ‘Save and compare’
Users would be able to save up to 4 cars without having an account, stored in a cookie.
To save more, users would be onboarded into a full account
Notification counts added to the ‘heart icon every time a user saved a car
Compare messaging would appear after a user saved more than 2 cars
Compare is initiated through the users saved adverts list
Users can compare up to 4 cars at any one time
Users would be given a overview and tech spec tabs with which to compare on
There would be a number of features we hoped to add in subsequent phases that had tested well with users
Versus mode with user selected categories
The ability to compare finance deals
Manufacturer sponsorship opportunities
The ability to ‘pin’ one car to a slot in the tool and the others are compared against that one
Desktop and app experience
First 4 listings are active and can be compared and saved
Outcomes and lessons
Launched to a 5% A/B test for a few weeks initial results were positive. During the first week we found
25% more users clicked ’Save and compare’ compared to ‘Save’
Adverts saved: 3,461 (test) vs 2,865 (control)
Access to saved adverts up 58%. Heart in navigation vs old link in users account
1,467 comparisons done
Average of 2.8 cars compared at a time
We rolled out o 100% eventually with a number of usability tweaks ready to go, identified following post launch user sessions. Unfortunately momentum on save and compare ground to a halt as resource was pulled onto other projects. Almost a year and a half later, to tie in with some consolidation of our navigations, the compare feature and saved advert lists were brought to the desktop experience.
Fast forward to 2020 and with a re-design of our car advert pages, the save action has become a little lost on the page and had lost the compare context to the action. We are in the process of making the entry into the users saved lists and the promotions of compare a lot stronger. Coming off lab study with a set of users this update is set to be tested on site in mid-march.
Compare updates 2020
