Google Case Study: Jumia harnesses machine learning to optimise bids while boosting conversions
November 20, 2019In the NewsResearch
Jumia is a pan-African e-commerce retailer offering millions of products from thousands of brands. The wide product range presented a problem, though: with so many accounts to manage, manual bidding was taking up significant time and resources. Jumia set out to reduce the amount of time the team was spending on optimisation tasks so they could focus on more important things like scaling categories and acquiring new customers.
The Smart Bidding exercise helped us understand our user behaviour better.
To improve the brand’s search presence and results at scale, Jumia started using Target CPA in their Google Ads account. Through this Smart Bidding strategy, bids were optimised automatically to help Jumia get as many conversions as possible, at or below the target cost per action (CPA) set by the team. Regular updates with visual reports showed the evolution of the bidding system and improvements to the results over time.
“Though we started on shaky grounds, the Smart Bidding exercise helped us understand our user behaviour better and we could see results getting better after a period of four to six weeks,” says Tabrez Firoz. “We noted an increase in conversions and better overall performance, while the effort spent on bid optimisation reduced.”
The reward of Jumia’s efforts are improved performance metrics: a 48% increase in conversion rate, 2.4% fall in cost per acquisition, and 52% growth in conversions, all with 33% growth in search impression share year over year.
Jumia is a leading e-commerce platform in Africa. Our marketplace is supported by our proprietary logistics business, Jumia Logistics, and our digital payment and fintech platform, JumiaPay. Jumia Logistics enables the seamless delivery of millions of packages while JumiaPay facilitates online payments and the distribution of a broad range of digital and financial services.
For more information about Jumia: Abdesslam Benzitouni [email protected]