Going Back in Time the CPG Way 
By Chris Enger

Time travel has featured in our culture in many forms, and scientists have long explored whether it might be achievable and controllable. What if we could make it a reality, and change the mistakes we have made?

For CPG executives in charge of large portfolios, “What if?” is more than an entertaining thought experiment. For them, real value is at stake. What if we increased the price of a specific product? Would customers keep buying it? And what if we ran a promotion for one of our brands? How would this affect other brands in our portfolio? What if we reduced the size of our assortment? Would the savings outweigh the losses?

In the past, companies have often relied on experience and intuition to guess the effects of the changes they make. But as portfolios grow, consumer decision making is evolving and commercial cycles are getting shorter. The experience of even the most seasoned experts cannot guarantee the right decisions. Most existing approaches are based on predicted consumer switching, rather than actual switching. In effect, the output of traditional models is neither as precise nor as reliable as it should be.

Same Problem, New Perspective
There is a new approach that we have been part of developing that does justice to the complexity of today’s portfolios and consumer decision journeys. The demand transfer matrix approach describes how volume flows in response to commercial decisions such as pricing changes, promotions and assortment restructuring. It is based on net elasticity of demand, rather than elasticity at the SKU level.

In other words, it reflects the total impact of a pricing or assortment change across a company’s entire portfolio of brands and products in the same category. It accounts for “good volume flow” within a company’s portfolio as well as for “bad volume flow” to competitor brands. In addition, the demand transfer matrix approach uses observed switching behavior as derived from household panel data to create a precise understanding of product substitutability, while it can also represent non-existing products through simulation.

One of the things this approach seeks to do is aim to combine the power of machine learning (ML) and artificial intelligence (AI) with more traditional approaches like prescriptive analytics to model, and, more significantly, recommend changes that can be made to assortment, pricing, promotions and SKUs.  Just providing teams with reams of data for further human analysis and debate stops short of enabling companies to realize the true benefits technology can bring to the bottom line.

Back in Time
Because it allows for simulation of changes in areas such as assortment, pricing and promotions, the demand transfer matrix approach is the next best thing to a time machine. In the future, it could be applied to maximize value creation across entire ecosystems of the economy. CPG companies could use the approach to establish a more fact-based incentivization of their retail partners. Why pay listing fees for a lighthouse item that ranks high in the consumer’s hierarchy of needs and attracts shoppers to the retailer’s stores? Conversely, higher listing fees might be required to get a retailer to stock items for which there are strong, substitutable alternatives.

Positive Impact
Although the demand transfer matrix approach is a relatively recent innovation, it can bring both strategic and tactical benefits for organizations, especially those managing multiple brands in the same category. Specific benefits of the approach include:

  • Combines the merits of hierarchical purchase structure analysis with econometric modelling at scale.
  • Helps companies analyze and simulate cross-elasticities and demand flows across large portfolios to calculate net volume, revenue and profit impact.
  • Disregards irrelevant switching, such as a customer buying a zucchini for dinner one day and an apple as a snack the next day.
  • Compares actual demand changes with predicted changes, rather than relying on predicted flows alone, to increase the accuracy of the output.
  • Predicts transfers even for low-volume SKUs and non-existent (new) items, while traditional models often neglect or underplay these effects.
  • Allows for differentiation of elasticities and preferences for different commercial levers, such as pricing, promotions and assortment changes.

Time for Change
The level of precision that can be brought to the decision-making process with the demand transfer matrix can give real confidence to CPG executives. It boils down the complexities of the many variables influencing the outcomes they desire into actionable insights. Going forward, the approach could even fuel the collaborative development of new SKUs or ranges to drive revenue, profit and loyalty.

Key to the success of this innovation is how companies approach the integration of AI and ML technologies into their decision-making processes. This is a critical component that will set leaders apart and allow them to outmanoeuvre competitors in terms of assortment, promotional activity and ultimately profit.  While we may never see the day that time travel becomes a reality for the human race, we can bring science to the table and influence the success of our CPG brands in the future, right now.

Chris Enger is senior expert at Periscope By McKinsey.

CPG VIEWPOINTS             
                                                                   Early December 2018
Tell a friend about this page