Intelligent decisioning in retail: Delivering on the personalisation promise


The retail sector has been in a state of change for many years now. Retailers have long been discussing the shift to online – or rather the correct balance between online and "bricks and mortar" – and how to cater to customers’ desire to use multiple channels for different parts of their journey. However, pandemic lockdowns around the world have starkly exposed this issue.

With nonessential bricks-and-mortar stores shut for an indefinite period, customers turned to online shopping. What’s more, they have been slow to return to stores as lockdowns were eased. Retailers everywhere have announced store closures and redundancies as they try to manage the fallout – or take advantage to make changes that had been needed for some time. But is this real change, or just final acceptance of the inevitable?

Colin Wright, Global Solutions Director of Retail and Consumer Goods at SAS partner Microsoft, suggested in a recent interview that there had been a paradigm shift in retail driven by the pandemic. He also commented that he believes the next few months and years will be hard for retail marketing, with budgets and head count greatly reduced.

However, this does not mean that everything is a disaster. Instead, it simply means that automation, and using analytics to gather intelligence, will become crucial in retailing.

Analytics in retail

There are several ways in which retailers can use analytics, and they all have one thing in common: They are about improving the customer experience.

The fundamental issue is forecasting demand. Early in the pandemic, we saw essential goods retailers like supermarkets caught on the back foot. Changes in buying behaviour, coupled with disturbances to supply chains, meant that they simply could not meet the demand for some products. Experts like my colleague Edward Kerrigan, Principal Client Adviser for EMEA Retail, suggest that retailers are likely to be taking a long, hard look at their forecasting models to build in more granularity. This will allow them to respond more rapidly and flexibly to changes in demand.

There are several ways in which retailers can use analytics, and they all have one thing in common: They are about improving the customer experience.

At the same time, demand for online grocery shopping increased massively – by 200% in some places and for some retailers. These retailers simply could not meet that demand. Many had to ask people to visit stores "if they were able" so that online slots could be reserved for more vulnerable customers. Similar spikes and troughs in demand were seen with other goods.

Approaches with analytics

One approach for responding to significant change is using analytics to allow short-term demand sensing to bring together customer needs and processes and then apply those to forecast demand. Machine learning can also draw on both current and historical data to improve prediction.

The critical challenge is how to predict and plan for substantial changes in consumer demand. SAS can help you improve demand sensing decisions across the supply chain: in the short term during the disease outbreak; in the midterm during the recovery period; and in the long term once the pandemic has subsided.

As the pandemic ebbs and flows around the world, there are some lessons that we can learn about forecasting demand in retail. Crucially, the most important aspect is likely to be keeping models up to date. This means rapid use of real-time or near-real-time data to drive operational decisions. When things are changing so fast, you cannot use a historical model. There are some patterns that have recurred in different countries, and they can be included in models. However, the most important thing is to respond rapidly to what is actually happening in your stores, and with your customers’ behaviour. Decisions must be driven by actual events, not an idealised prediction of what might happen based on data from five years ago.

Managing personalisation

One very interesting phenomenon that emerged during lockdowns was retailers’ inability to deliver for customers. People who had shopped for groceries online for years suddenly found that they were unable to get a delivery slot. Marketing departments resorted to emails from CEOs asking everyone to shop in store if possible. It was a long way from the individual personalisation that has become expected in retail.

Are we moving away from the expectation of personal service? I don’t think so. Instead, I think retailers have upped their game. They have put in place systems for better demand management, and better decision making, so they can deliver for customers in the future. Intelligent decisioning, supported by advanced analytics, is the key to meeting and exceeding customer needs in retail, whether in store or online. This is not about to change in the near future, pandemic or no pandemic.


About Author

Antonio Calvo

Industrial engineer from Catalonian Politecnical University UPC and MBA from IESE. With more than 20 years experience in international companies in EMEA in the Retail and CPG. I'm passionate about working closely with customers and executives in solving complex business problems, leveraging customer insights out of Big Data, advanced analytics and AI to provide and increased accuracy in strategic and operational recommendations and decisions. Large experience in developing and implementing projects in international companies from retail and CPG. Currently leading the business development for Retail and CPG industry in South EMEA region.

1 Comment

  1. Nancy Rudolph on

    Great post Antonio! The emphasis placed on retailers to "up their game" in personalization continues to play a large role for competitive advantage. Intelligent Decisioning is the key to making it all work!

Leave A Reply

Back to Top