Magalu Continues As An Ecommerce Pioneer With Deep Learning Ad Campaigns

Magazine Luiza - affectionately known as Magalu - was founded in Brazil in 1957. The company was an early pioneer of ecommerce solutions and has now evolved into one of the largest retail sites in Brazil, selling goods ranging from home appliances to cosmetics to sporting goods. In addition to more than 1,300 stores in 21 states, Magalu has six other online brands (including Netshoes, Zattini, and Época Cosméticos), 200,000 sellers in its marketplace, and the AiQFome food delivery platform. The company’s SuperApp has more than 45 M active users.

As a well-established ecommerce site, Magalu was already using retargeting solutions and needed additional partners that could increase revenue without duplicating existing customer touchpoints from other channels. RTB House needed to prove the incremental value of its campaign and promote healthy online growth. A further challenge was delivering these results at speed and scale. Due to the proximity to the Black Friday period, it was vital to efficiently scale up campaigns at key moments.

This initial campaign covered both app and web users between June and October. The Deep Learning-powered technology that underpins RTB House campaigns was able to identify new qualified users who had not been reached by other ad partners. These users were then shown more personalized ads with relevant product offers, as ultra-powerful Deep Learning algorithms can ascertain the interests of each individual based on their on-site behavior. This creates a win-win for Magalu and its customers: the brand is better able to reach its media goals, while customers get more personalized interaction with the brand and a better online shopping experience.

The speed at which Deep Learning algorithms are able to analyze behavior and autonomously calibrate campaigns to optimize efficiency allowed RTB House to achieve fast results for Magalu. As the campaign scaled up, we were able to maintain this effectiveness, bringing 3.2 times more revenue, 2.7 times more conversions, and 20% higher average order values (AOV) in October than in the opening month of the campaign. We hit monthly ROAS targets - including ahead of the crucial Black Friday period - proving that even a well-established ecommerce site can make significant improvements with the right ads. More importantly, we did so without diminishing the impact of existing marketing channels.