Blog

Do you know what your Customers will buy tomorrow? We do

As Retailers continue to increase their investment into data science capabilities, they’re quickly reaching the point at which they can make highly accurate predictions about how shoppers will behave in the future. In this article, dunnhumby’s Vijay Balaji Madheswaran asks what that really means – and how grocery Retailers can turn foresight into real competitive advantage.

From palm reading to astrology, for as long as we’ve been capable of cognitive thought humans have been obsessed with trying to predict the future. Who wouldn’t, after all, like to have advance warning of next week’s winning lottery numbers, or to be able to tell exactly how a decision will pan out before they’ve even made it?

The idea that we could somehow predict what will happen tomorrow is inherently captivating, borne out by the popularity of blockbuster films like Back to the Future and Minority Report. While time-travelling DeLoreans may be some way off yet though, continued advancement in our ability to process and understand data at gigantic scale means that we’re living in an age in which anticipating the future is a very real possibility.

In grocery Retail, data science now gives us the ability to predict shopper behaviours incredibly accurately. With a very high degree of certainty, we can say what someone will buy in a grocery store – and potentially even which store – during the next four weeks. That’s not some future ambition, or some false promise; it’s something that dunnhumby helps Retailers around the world with every single day, using a combination of purchasing and loyalty data alongside sophisticated predictive algorithms.

The logical follow up to that statement, of course, is “so what?”. What does that foresight really enable us to do?

I think that the answer to that question becomes clearer when we look at the typical responsibilities of a Chief Marketing Officer (CMO) within the grocery industry. While specific circumstances will vary from company to company, it’s not unreasonable to suggest that most CMOs are tasked primarily with:

  • Helping the business acquire new Customers
  • Maintaining the loyalty of existing Customers
  • Maximising the value of every shopper

Data science – and by extension, behavioural prediction – can have a huge impact on how CMOs deliver on those objectives.

  • Identification and acquisition
    From a data science perspective, acquisition is a little trickier in that Retailers obviously don’t have insight into people who aren’t yet shopping with them. At the same time, third-party data can fill in many of those blanks. More importantly, perhaps, insight into existing shoppers can be hugely useful in terms of helping Retailers understand the “right” kinds of Customers to acquire next.
    Segmentation – the process of breaking down your shopper base into smaller groups – allows you to categorise shoppers in respect to the value they bring to your business. By knowing how high-value shoppers behave, and what their needs are, Retailers can gain a better understanding of what they need to do in order to attract more of them.
  • Rewarding shoppers and maintaining loyalty
    What is it that really keeps people loyal to a store? Why have shoppers behaved in certain ways in the past? What has caused Customers to stop shopping at a store? Data science can answer all of those questions, giving Retailers the insights they need to create high-level engagement strategies that help Customers to feel like they’re getting better value for money.

    • That might mean delivering on better promotions, or it could be about refining assortment; the important thing is that by looking to the past, Retailers have a much clearer idea about how they need to respond in the future.
  • Maximising value through upsell and cross-sell
    The vast majority of grocery Retail Customers almost certainly don’t spend as much as they could with a store, meaning that share of wallet is still there to be gained. In my experience, that usually comes down to issues with range; either shoppers can’t find the products that they’re looking for, or the store isn’t located in a convenient location for their current shopping “mission”.

    • Ensuring that the right selection of products is available for the right shoppers in the right locations is something that data science excels at, as is its ability to aid in identifying which items to highlight as part of Retail Media campaigns. If you know for a fact that the problem isn’t with your range itself, but with shopper awareness of it, data-driven Retail Media offers a highly effective way to raise awareness and influence purchasing decisions.

The common theme across all three of these areas is that data provides the ability to experiment, test, and learn. By making minor adjustments to their offering and trialling them with small subsets of shoppers, Retailers can predict how their future decisions will play out before they roll them out to a wider audience.

The stark reality of course, is that not every Retail organisation is currently set up to handle data in this way. The fear of Customer information getting lost or compromised is a longstanding one, and an issue that carries increasingly harsh penalties in many regions. Moreover, while there might be a growing fixation with “real-time” decision making in Retail, most grocery businesses just don’t have the operational infrastructure required to deliver on that – technical debt and legacy systems standing in the way.

Nonetheless, while it might not be simple, the opportunity to use predictive science to build a better and more sustainable grocery business is there. As gains in data literacy and automation provide data scientists with the space they need to deliver even greater value to the organisations they work for, advances in AI and Machine Learning will only aid them further. Retailers will be able to make smarter and faster decisions about how to win and maintain Customers than ever before.

Best of all, by letting the data guide us, we can do so in a way that’s better for those shoppers. It tells us everything we need to know about how to inspire them, engage them, meet their needs, and delight them. And if we can build lasting value for the shopper, we can build lasting value for the Retailer too.

Vijay Balaji Madheswaran is Director of Applied Data Science for dunnhumby APAC. Focused on investment, partnership, access, and customization, Vijay is passionate about helping Retailers and Brands realize the full value of their data.

 

The latest insights from our experts around the world

Understanding the why: Causal AI and its value for retail
EBIT: the next milestone in retail media measurement?
Where will our data live tomorrow?
customer first data science analytics & machine learning services
Ready to get started?

Speak to a member of our team for more information