MOTORS

The dealer information and vehicle inventory were stored in separate places, so we needed to combine this data using Google Cloud and BigQuery. Dealer IDs were especially tricky because we had to bring all the related data together. Once everything was unified, we could create one complete feed, which was then formatted to work with platforms like Google and Meta for the ads.

Due to the size of Motors’ inventory, data impurities were inevitable. Some sold or removed vehicles remained in the feed with links to inactive pages. We identified out-of-stock vehicles through image links showing a default ‘no image found’ picture. To fix this, we set up filtration rules to remove outdated vehicles, exclude outdated images, and eliminate flagged errors from the Google Merchant Center’s diagnostics report.

We wanted to prioritise advertising for high-demand vehicles, so they received more budget. To do this, we created a dynamic hierarchy in the feed, aligning bidding strategies with Motors’ goals. We developed a labelling system that classified vehicles and dealers into tiers, allowing us to adjust bids and focus on high-value vehicles for better campaign efficiency.

Google Vehicle Ads (GVAs) was a Beta ad format using a new feed structure, separate from traditional product feeds. We collaborated closely with Google to understand the feed’s parameters, creating a solution that met the minimum requirements and added extra features to maximise the format and gain a competitive edge.

This involved mapping vehicle attributes like make, model, year, and body type into Google’s required fields alongside each dealership as a store location, linking vehicles to their respective dealers. The feed was also transformed into a Meta-compliant version, which meant we were able to power dynamic ads across both platforms.