New car: search optimisation
Overview
With over 40 years of heritage Auto Trader has always been known for being the place to go for a used car. As the car buying landscape evolves, new cars are becoming more affordable through cheap monthly finance deals and new ownership schemes like PCP (Personal Contract Plans) and PCH (Personal Contract Hire). As the new car squad, we had been responsible for the front end presentation and journeys for bringing brand new cars on to Auto Trader. Working with our sales and trade functions we were able to add unregistered, dealership allocated, brand new cars into the core Auto Trader search journey.
In fields and warehouses across the country, there are thousands of these brand new cars ready to be sold, but no way for the car buying public to access them apart from a spreadsheet on a sales managers computer. After many iterations and a few pivots we had a compelling new car proposition we hoped would create value for our dealership customers and bring transparency to pricing by displaying pre-haggled prices for our consumers.
With a new cars live on site and starting to gain some momentum, it was time to start optimising the product for conversion with a view towards monetisation.
Problem statement
With monetisation of the new car product approaching, we needed to optimise the new car product to increase conversion from search and generate more leads sent by consumers to our dealer customers.
Users and audience
We know from Auto Trader research that around 70% of the car buying population are open to either a new car or used car, with price and transparency of pricing, playing a major part in any decision.
Roles and responsibilities
Lead product designer for the new car squad. Responsible for all new car touch points throughout Auto Trader from discovery, search and contact with a seller. 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
Chunked up to 6 week slots, constraints were set around the objectives from the OKRs we had created for the new car project.
Objective 1:
Increase the visibility of new car on AutoTrader
Key Result 1
Increase new car search impressions per car per day from 125 to 156
Key Result 2
Increase new car FPA views per car per day from 3 to 5
Objective 2:
Improve the performance of new car adverts on AutoTrader
Key Result 1
Increase new car CTR from listings to FPA from 2.4 to 4
Key Result 2
Increase new car conversion from FPA views to lead from 0.40% to 2%
Of these two objectives, it was second was one we had the most control over. We would be measured on the first one, but this need much more collaboration with other squads / teams within Auto Trader.
Our process
As a new car squad, we had a passionate, engaged team and this optimisation project was a perfect opportunity to encourage a sense of ownership of ideas and solutions. The best teams think and ideate together. To me, it is the job of a product designer to be a facilitator of ideas, realise their potential and using systemic thinking and rapid experimentation, deliver measurable outcomes for our key objectives.
OKR Workshop 1
The first of the 8 week chunks were to be based around the key result of ’Increasing new car conversion from FPA views to lead’.
With an objective and result set, it was time for the first design task, to create a workshop that can best facilitate the ideas of the team. With research (internal & external) and data underpinning the day and a strong theme of ideation and collaboration we were set to kick off the workshop.
The day started with an introduction to the overall new car team OKRs and the benchmarks we had set against those. Working with our internal research function we were able to paint a picture of current usage and performance of the new car product. Engagement from the team had started early and was encouraging to see note taking during the research session.
As product teams it’s important to be aware of how other companies approach similar problems. With that in mind the team had been set some homework to present on the day, ‘Find a conversion experience on a non car related site that you find easy and delightful. Prepare a couple of slides on what made that experience delightful and easy to use. ‘ It was good to see the variety of examples the team brought to the workshop and they sparked a number of themes that would last across the day.
From here the team were asked to come up with some ‘How Might We’ statements, free of any constraints or current Auto Trader patterns. These were high level broad themes that provided enough detail to make sense but didn’t stray into solutions. It’s important to encourage the wackier and crazier end of the ideas spectrum as it is much easier to reign in a crazy idea than it is to make a mediocre idea interesting.
A huge wall of post it notes were collated into various similar ideas. Using silent dot voting we swarmed around 2 of these themes. The first of them focused around the presentation of the advert / product page and the second focussed around the seller contact form. The team split into two to focus on these themes and through various individual and team sketching sessions and countless iterations and discussions, the team came back to present their ideas.
Armed with pages of ideas and working through an effort vs impact matrix, it was my job now to realise the potential of these ideas and formulate a prioritised testing plan.
Pricing on advert page
Presentation of the pre-haggled pricing.
Setting a ‘Was, Now, Save’ format to the pricing and the visual execution of that. Through a series of small quick tests we arrived at the pre-haggled price in red combined with strikethrough execution of the ‘Was’ price and the monetary saving in aqua. Further tests removed the ‘Now’ copy and changed the ‘Was’ pricing to ‘RRP’.
Price and contact action alignment
Alignment of pricing and main call to actions
Through the workshop, there was strong theme of the close alignment of pricing and calls to action. Through various quick design tests and live A/B tests we arrived at a theory we could organise the page around. Utilising e-commerce patterns, we arranged the ‘what’ (make and model of the car), with the ‘how much’ (visual presentation of pricing and saving) and the ‘how to’ (main deal contact call to actions), much more closely together than the current used car execution of the page.
New alignment resulted in an increase in click through rate of 4%
New car form tests
Simplification of the contact form
Armed with usage data of the current new and used contact forms we were able, through quick fire tests to start chipping away at the construction of the form, to the what we felt was the simplest but still valuable execution. Form elements were taken out, added back in and every iteration in between. We got down to a form where we had removed navigation and other distractions from the page to reduce the visual cognitive load, removed the pre selected questions and replaced that with a free text message box, made the phone number optional, removed the marketing opt in to the next step and consolidated the name fields down to one field.
Form completion rate increased by 3.5%
OKR Workshop 2
The second of the 8 week chunks, were still based around the key result of ’Increase new car CTR from listings to FPA’.
As the workshop worked well last time, we retained the format, adding in an update on the OKRs from the last workshop.
Again with a stack full of great ideas and a view of prioritised ideas based on effort vs impact, we set about testing and validating these ideas.
Pre-haggled pricing in search
Pre-haggled pricing in search
Building off the success of the pricing executions in the advert page, we tested again but this time through search listings
A listing showing a discount is 2.5% more likely to be clicked. When combined with a actual photo of the car, this rises to 8%
New car persuasion messages
Persuasion messages and listings executions.
Using search listings to highlight the differentiators with new cars such as warranties, copy to indicate a ‘deal’ e.g. ‘Special Offer and showing the percentage saving.
Persuasion messages decreased click through by 1.5%
Deals toggle ideas
Deals toggle in search
Launched within the search form, a toggle that would display in search results, only those new cars that had pre-haggled prices associated to it, a ‘deal’. A member of the tam had brought the Amazon Prime toggle as part of their homework, so this based the thinking behind these tests.
New car only searches saw an increase
Outcomes and lessons
The biggest takeaway from these projects was seeing the sense of ownership the team took, when they were empowered, to ideate and see these ideas through to live executions which delivered measurable results for the business. New car continues to grow from strength to strength within Auto Trader, through February 2020 we now have;
Over 32,000 brand new, unregistered cars for sale
5 million new car advert page views for February, up 14% month on month
66% of new car stock displays the pre-haggled pricing
New cars with pre-haggled pricing generates 2.5% more click through than standard new car adverts
Over 13,000 new car leads delivered in February, up 21% month on month
8.4% of users seeing a brand new car
Nearly 1,000 paying customers on the new car product
