For as long as consumer packaged goods companies (CPGs) have been using technology, they have been criticised for not using the latest or the right tech, to the extent that many have become inured to the negativity. Even feedback that takes a more positive approach generally comes with a sting in the tail, that what is working well now cannot continue to succeed in a fast-changing market.
Of course, almost all CPGs want to change, to a greater or lesser extent, not least more lately in a market that has seen more changes in five years than it had in the 20 before. The constraints, of course, were always the same – time, money, culture and the potential impact of disruption caused by new tech.
It is now fair to say that these reasons can no longer be put forward as an excuse not to act. In the face of a health crisis that has accelerated many of the changes happening in the consumer industries for the last five years, everyone’s idea of normal will continue to be upended for the next year at least.
Demand signals at the retail shelf are all over the place. There is no sign they are settling, and new types of demand behaviours are emerging that have never been seen before. Just one example is the growing gap between what consumers will pay for a basket of goods. They are faced with the conflict of having to manage between not caring about the price of certain essentials while, at the same time, looking for the best possible price because they are worried about whether they will have a job soon.
And basket size and mix has changed, which affects the traditional relationships between categories as many households will shift to value for certain staple products and trade up to new ones to treat themselves to replace their lost visits to restaurants.
Measuring customer behaviour and sentiment inside that paradox is impossible.
However, while it is possible to create highly accurate demand forecasts as we will show, the way most CPGs are structured will inhibit their access to them. First of all, accurate demand forecasting depends on multiple sources of data, both internal and external – product lifecycle, seasonality, trends, market entry, holiday and other peak demand curves, promotions and price changes. It also requires greater collaboration with retail partners, not just in the category but across categories.
But because ownership of data is spread across departments, a single version of the truth is not available, and therefore cannot be exploited by departments that do not collaborate around shared goals. This is the first hurdle.
Once departments agree to collaborate around a shared view of data in order to build a unified demand strategy, the next hurdle is to extract meaning from the multiple sources which is not possible with human processing on spreadsheets. And in a global pandemic, it is even harder for CPGs to create the ideal plan for every product, category and channel. Covid-19 buying behaviours and emerging demand patterns don’t follow any historical data models, so planners can no longer lean on the past to predict the future.
The solution is to use artificial intelligence to smartly integrate these data sets to create a unified demand signal that generates plans for all categories. Several models can then be created to exploit particular opportunities that follow retailer partners’ own goals; to manage data challenging products such as short life cycle, new or end-of-life products; to help them fix categories that are not performing well; or to help find growth opportunities by expanding inventory or changing assortment to gain sales.
In no way is the role of the category manager diminished. Aside from the obvious superhero status they can assume by improving volumes and/or sales price, they can rise to the more exciting role of pioneer, creating scenarios that may reveal new products and opportunities and which can be tested before being rolled out.
Hopes for a new normal built on top of the old normal are being dashed daily as more bad news is announced around the impact of covid-19. More demand will shift from planned to unplanned, and only data science can equip the category manager with the tools necessary to create a unified demand strategy.