Businesses have relied on data to make decisions throughout history and in today’s dynamic business climate, data-driven decisions are more important than ever, to empower data-driven approaches.. Fortunately, advanced technology provides e-commerce businesses with a distinct advantage, but the reality is that we still have a way to go in terms of harnessing the full power of Big Data.
There are three key challenges facing e-commerce today. Firstly, the volume of valuable data generated by existing and potential online customers is immense. Secondly, many of the available cost-effective tools can’t process this data at a rate that supports timely, and in some cases real-time, reactions. Thirdly, there is a considerable knowledge gap amongst the pool of human resources needed for such efforts. As a result, an e-commerce marketplace may struggle to convert this data into real value for various stakeholders.
Applying Big Data techniques
In order to meet these challenges head-on, Jumia’s first order of business was to put together a core tech team in charge of centralizing the disparate data sets generated during our day-to-day operations. Approximately two and a half years ago, this team started creating a big data platform underpinned by a data lake, to leverage analytics and data science initiatives across Jumia.
Soon thereafter, we discovered that we also needed to apply Big Data techniques since we required a top-to-bottom approach with complementary cultural shifts. This is why we then deployed a team of data engineers to serve as conduits between the core tech team and the numerous functional teams.
In practice, the tech team would set the primary direction of our Big Data efforts while data engineers would enlighten other teams on contributing to these efforts. These engineers worked with marketing, commercial, logistics and other teams to ensure that their existing tools and processes could feed the central Big Data platform consistently.
Where necessary, an engineer liaises with a specific team to refine any procedures related to collecting relevant data points so that they are in harmony with the tech team’s objectives.
Big Data at work
There are plenty of ways in which we use Big Data to enhance the e-commerce experience for both sellers and consumers, and so far:
Our team has managed to run informed data science projects that result in more intelligent platforms. E.g., Jumia can offer more precise delivery times based on multiple factors and even predict the ideal routes that delivery agents can rely on to meet their targets. Consequently, consumers enjoy a better experience since they are more certain about the progress of their orders from purchasing to delivery.
Sellers can access data spanning various metrics through the seller center. These resources can help them get a broader view of their stores’ potency aside from their revenue. They’ll have a clearer picture of how consumers interact with their product offers, the stages at which they lose conversions, perceptions surrounding their brands, and many other insights. From here, they can make more informed decisions on how to position themselves on our platform and gauge the results of their efforts in near real-time.
We can apply Artificial Intelligence (AI) and its derivatives such as machine learning to make more resonant product recommendations. When making suggestions, the data science team can create and refine algorithms that consider price, browsing history, previous purchases, and other data points available in the data lake.
Ultimately, consumers enjoy the shortest route to the next best alternative, whether it’s through search suggestions, related product sections, category shortcuts and more. In essence, they feel like the platform deeply understands them and what they are trying to achieve when shopping.
The central Big Data platform creates opportunities for different service providers to plug in their Application Programming Interfaces (APIs) and synchronize user-facing tools with back-end functions. For example, platform users will have an easier time extracting and utilizing account information needed for tax purposes. Additionally, the integration capabilities make functions like fraud detection more seamlessly embedded within the e-commerce experience
For every initiative we undertake to grow e-commerce in Africa, we strive to do more than achieve accuracy within internal operations. Using Big Data and supplementary techniques like AI, we can go beyond deriving actionable insights from multiple sources.
Our teams develop ultramodern features and functions that encapsulate real value for consumers and sellers alike. The goal is to make it simpler and safer for people to buy or sell goods through our platform, irrespective of any prevailing impediments both online and on the ground. ***********************************************
Hélder Quintela Big Data Manager at Jumia PTC Hélder Quintela is the Business Intelligence and Big Data Manager at Jumia. Hélder joined Jumia in 2019, working in the Jumia Porto Tech Center. With a degree in Master Science Information Systems, Hélder has a strong experience in the data field, in areas such as Healthcare, Finance, Education and Industry.
He is currently Vice-President of the Scientific Council of APBI, the Portuguese Association of Business Intelligence and an invited teacher in the Polytechnic Institute of Cávado and Ave - Portugal. As BI & Big Data Manager he is one of the people responsible for the implementation of the strategy for usage of data in Jumia.