Demystifying data science and looking to the future through the AI lens
What’s trending and will be the next must-have for any merchant? Naturally, the latest technologies used to take businesses to the next level, helping them scale. To dive further into this topic, we sat down with João Moura, Head of Data Science at Payvision, to discuss two intertwined topics that are becoming the subject of every e-commerce conversation. Payvision is a data-driven omnichannel payment provider that makes payments simple for merchants around the globe, supporting them in growing their business.
Ecommerce Foundation (EF): Before we get into detail about these two topics, let’s start with the basics: How is data science adding value to merchants?
João Moura (JM): Keeping up with modern-day consumers requires merchants and retailers to constantly monitor the journey of their consumers. In order to do that, they make use of data science – crunching the numbers and extracting valuable consumer journey knowledge that they can later translate into business decisions and strategies. Data science means companies now have the power to use customer intelligence insights to take their businesses to the next level. Knowing your customers’ preferences is the first step in becoming truly customer-centric.
EF: Indeed, customer centricity is all about knowing your customers very well and taking that into account with regards to business decisions. So how can merchants build a clear data strategy?
JM: They should start by creating a clear and simple data architecture. This works as a base for how the business will access, manage and utilize data in a scalable manner. Everything starts with a landscape of various data sources, followed by mapping this into a centralized, well-governed data lake that will help merchants leverage their data in an effective way.
A clear data architecture is essential from the very start in order to get cleaned, defined data. In time, the data will improve in quality and accuracy and merchants will be able to avoid inaccuracies due to any conflicting information.
Next to implementing a clear data architecture, there’s another must that merchants need to be mindful of. A data strategy needs to be developed together with other departments so that it aligns with the overarching business needs. It’s crucial that they understand the benefits, otherwise data science efforts will encounter roadblocks right from the start. This is very important for merchants to keep in mind: get involvement, alignment and approval from all stakeholders from the very beginning.
In this sense, it’s crucial to have an internal awareness campaign to inform staff about what the company can do with data so they understand the benefits. Unless employees are familiar with the advantages and see the value, data science initiatives will not work. In this sense, merchants need to make sure they have a game plan and start this awareness campaign from the very beginning. They might even want to get help from their internal communications team who can support them with the campaign. Together, they need to come up with different activities to keep staff engaged in this campaign.
EF: Right, so that’s the part that merchants need to sort out first. It obviously makes sense, but how can they enable data science to be a game changer for their business?
JM: There’s really no secret recipe; merchants need to take data science seriously from the start.
By doing this, they will avoid their data science initiatives ending up as simply R&D projects. Reaching this stage makes it much harder to produce meaningful results. Data science models and artificial intelligence (AI) must be built into processes, applications and dashboards so that the benefits are easily accessible for end users. For this, merchants need to dedicate the right planning and resources. In turn, employees will start to see the advantages of data science initiatives and then trust their potential. Next to that, it’s absolutely essential that the data science team has a good understanding of the business itself so that they can identify the potential benefits and translate them into different initiatives. Maybe it sounds strange, but the implementation of technology can be much easier than actually understanding the data potential. That’s why it’s very important that merchants support their data science team so that they understand as much as possible about the industry and business. This way, they have a holistic view and can help the company grow.
EF: But is there a way merchants can increase the chance that real value is gained from data science activities?
JM: There are two levels of data science initiatives. Businesses usually start by being insights-driven, using knowledge from Business Intelligence techniques and then move on to the next level. The second phase is about becoming data-driven, where AI models are at the heart of the decision-making process, with all procedures automatized. Being insights-driven is a great way for merchants to start their data science initiatives; that’s basically the first step towards growing their business with data. However, it’s very important that merchants don’t get stuck in the insights-driven phase – they need to move on to becoming data-driven. This will help businesses see the data science results on a much larger scale while speeding up the process.
EF: Since we learned a bit about the data science basics, let’s take a look at the future and see what exactly it holds for artificial intelligence. It is said that in no more than ten years, the online payments industry will be almost entirely AI-driven. How will this impact the payments industry?
JM: That’s an easy assumption to make because everything will be AI-driven in ten years. In this sense, all business operations related to fraud, risk, and optimization processes will be powered and enhanced by AI. It will be a complete transformation of businesses as we know them today. And the future looks good: there will be a lot of automation, allowing professionals to have more time to focus on complex scenarios that require human work in instances where the machine can no longer dive deeper. We can also expect to see a change in dynamics: in time the AI automated services could reach a low-cost level, whereas the human-facing services could become the new premium standard. Automation is a natural step forward that will enable every part of a business to scale.
EF: That’s very interesting and we’re curious to see how that’s going to evolve over the years. Let’s also look at regulation and compliance. Do you think that the implementation of AI will be as successful in China and the United States with the European data policies being implemented?
JM: The population’s trust will be higher than in China and the US, which will potentially raise adoption. I believe regulations like GDPR will empower data science. With regulations in place, consumers have the guarantee that their privacy is respected, and they automatically feel safer about their personal data. At the same time, data scientists are also more comfortable in their work. We could say that they sleep better at night because with data protection guidelines, they have regulated access to data. The data regulations don’t hinder any potential in data science initiatives. Professionals can continue their work just like they did before. The only difference is that they don’t have to worry about being exposed to sensitive consumer data. They can still analyze data and determine patterns in the consumer journey, with all the necessary regulations applied.