The situation facing the car industry as a whole was a general trend towards delaying large purchase decisions. This lead to a reliance on money-off promotions to encourage purchases.
My design team used data driven insights to identify opportunities specific to Mercedes-Benz Online Showroom where we could defy this trend. We knew the sales performance of all Mercedes-Benz cars and had access to data from existing customers who generated a lead on one car, who ultimately purchased a higher priced car and so on.
With all this in mind, I set some boundaries for my team: Because this problem was open ended, any ideas would need to be flexible, with minimal technical debt in order to lower risk and allow room to explore insights as we discovered them. I wanted to make sure we didn’t interrupt any lead generating journeys in this process and finally, the challenge I set was: Can we do this without requiring money-off promotions or extra marketing spend?
Customers searching for Mercedes-Benz cars were in some way, already defying assumptions. Our hypothesis was that they could be ‘susceptible to suggestion’ in the form of up-selling to higher priced cars at various points in their journeys.
We extended the results card component from our exisitng design system to include a new ‘cross-sell tile’. This behaved as an instance of the results card object, giving us the flexibility in testing we were looking for. The cross-sell tile when tapped, would modify filter selections to display higher priced options on the results page.
Careful consideration was taken for the content of the first A/B test: We chose searches related to the relatively affordable A-Class and used the cross-sell tile to up-sell those A-Class customers to the A35 AMG (the performance variation of the same car). Both cars were being sold at the full retail price for the duration of the test.
A subset of customers interacted with the cross-sell tile and generated leads on the A35 AMG. We were able to increase month-on-month leads for the the entire A-Class range. We also ran an several variations of this test with other cars including petrol-to-electric cars and smaller-to-larger SUVs.
We were able to increase leads for the cars featured in the cross-sell tile. More interestingly, we increased leads on the original cars customer searched for as well.
When comparing the journeys of customers who were shown a cross-sell tile vs those who weren’t, the existence of the cross-sell tile appeared to encourage some customers to double-down on their original car selection. They generated leads more quickly, and in higher volumes, when the cross-sell tile was present, even when they didn't interact with it.
Over time we used this cross-sell process more widely to gernerate insights
into customer behaviour in support of multiple business activities related
to the
- MoM product page views (KPI) for the original car increased between 3.4% and 14.3%
- MoM product page views (KPI) for the up-sold car increased between 14.2% and 101%
- MoM leads generated for all featured cars increased between 1.82% and 163%
- We demonstrated to the business that we could increase leads without using money-off promotions
- Brand trust allows a level of influence in customers purchase decisions
- It is possible to up-sell customers to higher priced cars and generate leads for the original car at the same time
- There are often opportunities to influence behaviour that subverts wider trends
—