AI for business as UNusual

AI for business

We understand how consumer behaviours have changed, now it is time to act on them rather than go back to the old way of things, says Sivakumar Lakshmanan, co-CEO of antuit.ai.

We have been seeing unusual in- and cross-channel shopping patterns and behaviours post pandemic, and there is no shortage of sources to demonstrate this. The first research to emerge was able to observe early and mid-pandemic behaviour and many people assumed that this would simply return to normal once lockdowns and other restrictions were eased. However, as our own research reveals, this is far from the case and the world we now live in is one of permanent shifts. The massive pivot to online, local retail, new channels and alternative fulfilment has changed consumer behaviour dramatically, and to some extent forever.

Consider the scale of the changes bearing in mind that the landscape is still moving and will neither stop moving nor settle back to the way it was. In some cases, already observed behaviours have changed faster than anyone had guessed and in others, new behaviours have emerged. According to McKinsey, over 60% of global consumers have changed shopping behaviour, many of them for convenience and value.

In the antuit.ai research, 53.1% of respondents said that they would start or continue to prioritise spend on groceries over going out for meals, while the same percentage said they would continue to buy more grocery items to cook and bake from scratch in preference to going out since the pandemic. And 46% said they would carry on with ‘one big shop’ post-pandemic rather than going back to ‘top-up’ shops, while 35% say they would go back to ‘top-up shops.’

In further research, new implications emerge for both retailers and their suppliers. Over a third of respondents said they had traded down from branded grocery goods to own-brand during lockdown, and plan to stick with the changes they had made, while only 16.2% said they had traded up from own-brand to branded grocery goods during lockdown, and planned plan to stick with the changes. In addition to changing brands, 60.6% said they had become more sensitive to pricing over the last year (2021) while 61.6% had become more aware / sensitive to promotions and offers.

And given that these changes are observable across both on and offline baskets, it is now becoming clear that current data sets used to manage products across the life cycle are no longer an accurate indicator of demand if retailers are to respond more quickly, accurately and with agility to these trends.

Historical data has lost a lot of its power to determine demand, not just in terms of volumes, channels and basket mix, but customer loyalty; for instance, new consumer cohorts have emerged with no knowledge of or connection to brands’ heritage, which will affect decisions on ranging, assortment, pricing and promotion. In addition, existing consumer segments have fragmented that may make some traditional segments redundant

These trends will also shine a spotlight on techniques that were already not working before Covid. Using current techniques for managing products across the life cycle will only result in bad decisions on inventory – too much in the wrong place, too little in fast moving categories, too much slow-moving inventory and potentially more waste, which is becoming a reputational problem for grocers committed to reducing it.

The problem continues right at the heart of the way retailers and brands collaborate. For example, Nielsen says that two thirds of promotions fail, confirmed by deeper analysis showing that between 80-90% don’t unleash additional revenue or margin.

Solving the problem depends not just on better data and process, but on the ability to put insights to work across department and teams, not piecemeal. In retail and CPG, the goal is the creation of a unified demand signal, that gives each discipline within the business what it needs to supercharge, not only forecasting, but your assortment, allocation, and pricing strategies to perform optimally.

The sheer complexity of the task is tailor made for artificial intelligence which can handle multiple data sets, both internal and third party, as well as insights that can be discovered by data science teams through building demand models that can embrace the key KPIs of ordering, price, promotion and replenishment.

The good news is that these compelling business cases are starting to come through and proof comes from both Forrester and Gartner. Retail Planning is a key category in Forrester’s ‘Tech Tide : AI And Analytics For Retail, Q2 2021’ while its Now Tech report lists key vendors in the space. Gartner talks about 23 use cases for AI in with price, promotion and markdown optimisation, and in-store on-shelf availability at the top of the list.

As the store tries to embrace multiple roles forced on it by the huge shift to online orders, AI will have an even more critical role to play to manage new data sets and new data behaviours as yet unseen, for which history is no longer the benchmark.

By Sivakumar Lakshmanan, co-CEO of antuit.ai

[class^="wpforms-"]
[class^="wpforms-"]
[class^="wpforms-"]
[class^="wpforms-"]
[class^="wpforms-"]
[class^="wpforms-"]
[class^="wpforms-"]
[class^="wpforms-"]