Cost, efficiency, and profit: Why it’s worth testing automation

Heleen Blom, Hanna Riehl / December 2019

Digital transformation, machine learning, automation — knowing the meaning of marketing buzzwords is one thing, but knowing what they look like in practice is quite another.

But rather than seeing this as daunting, it encouraged Dutch online retailer vidaXL to investigate how they would utilise automation to make their digital advertising more efficient.

“Our success is based on the belief that things can always be better and more affordable,” says Stefan Deriks, Search Engine Advertising Specialist at vidaXL, and also responsible for performance display campaigns. “With growth as our key objective, we needed to find a way to reach customers at all stages in the buying cycle, rather than just those actively searching for furniture or any of our other categories.”

In trying out different automated solutions, vidaXL significantly increased their return on ad spend and boosted the customer experience and, subsequently, their own performance. Here’s how they did it.

Finding new customers...

To build awareness and cultivate intent among browsing users, the retailer placed display ads on websites established as most relevant based on a range of signals. “The display network is a great initial touchpoint, as it allows you to introduce yourself to potential customers,” Deriks explains.

“The only way to keep up with ever-evolving consumer needs is by continuously gathering and analysing data,” states Deriks, “which is where automation can really make a difference.” Using machine learning, Google’s Smart Display campaigns enabled the team at vidaXL to automatically connect with customers likely to be interested in their products, create ads that fit anywhere across the Google Display Network, and set the right bids to meet the advertiser’s performance goals.

…and staying in touch with existing ones

While finding new customers was a good first step, vidaXL knew that keeping them engaged was equally important. If a customer had viewed, say, a desk chair on the retailer’s site, they would use dynamic remarketing to create tailored ads including images of that same chair. By supplying a feed of product information — from price to materials and stock availability — and activating optimised bidding, they let automation do the heavy lifting in serving up personalised ads at the most efficient cost.

Cost, efficiency, and profit: Why it’s worth testing automation

“This approach helped us to reinforce brand awareness and remind audiences to make a purchase at a crucial moment in the user journey,” Deriks explains. As a result, the return on ad spend (ROAS) and overall performance significantly improved. Dynamic remarketing helped the conversion rate increase by more than 48% and improved click-through-rates by 24%, while the average cost-per-click dropped by 39%.

Scaling across the globe

Initially just set up for testing in the Netherlands, the campaign's great results prompted vidaXL to scale across all markets they operated in, from Europe to the US and Australia. “We have now implemented Smart Bidding in over 29 countries, with overall performances being good and steady,” says Deriks.

Although they had already A/B tested Smart Bidding across other Google channels, including Search, adding it to their Smart Display and remarketing campaigns allowed the team at vidaXL to effectively build on the data they had collected, and boost results results even further. Through machine learning, they were able to set optimal bids based on ROAS, and keep costs down as much as possible.

Driven by a desire to continuously improve their performance and cost efficiency, vidaXL saw automation for what it really is: a tool to help marketers reach their goals more effectively, so that they can stay relevant in today’s rapidly changing digital landscape.

Google Marketing Live 2019: Industry leaders share their key takeaways