Are you ready for a paradigm shift?

The rise of e-commerce calls for new digital demand-planning techniques in consumer and retail

Matt Kramer

Matt Kramer

National Sector Leader, Consumer & Retail, KPMG LLP

+1 614-241-4666

Scott Rankin

Scott Rankin

Principal, Deal Advisory and Strategy, KPMG LLP

+1 617-988-1474

Ken Spigarelli

Ken Spigarelli

Advisory Managing Director/Supply Chain Consumer and Retail Lead, KPMG US

+1 734-231-1311

Midway through the second year of COVID-19, consumer and retail executives are still closely monitoring the implications of the virus and variants, but have been showing signs of optimism about the future, evident in increased consumer spending and less price discounting. In the KPMG 2021 Consumer & Retail CEO Outlook Pulse Survey, a large majority of chief executive officers (CEOs) said they have confidence in the longer-term growth prospects for consumer and retail businesses globally (92 percent) and their own companies specifically (90 percent).

A strong vaccination program and the fiscal and monetary firewall built around the economy have supported rising consumer demand across multiple sectors. Consumer behavior, market trends, and preferences are impacting all participants. As an example, during the pandemic, we saw food and beverage demand surge in grocery retail while food service demand was all but non-existent. This trend has now shifted as some consumers head back to the office and start to frequent restaurants again. All of which puts an increased emphasis on meaningful demand planning and increased data analytics to predict volumes and drive better insights as consumer behaviors continue to shift.

Many of the changes we have seen during the past year—from buying behaviors, workforce models, and channel mixes—were well underway before COVID-19. Consider the shift in online buying and curbside pickup as two examples. The pandemic accelerated these trends, and forced executives to rethink fundamental assumptions, such as how to predict consumer demand by channel and how this demand impacted the broader supply chain.

The pandemic will continue to have an impact on the global economy in the immediate future. Demand and supply shocks will continue to drive volatility in the market and influence the sustainability of supply chains. We were not surprised when more than one-third of consumer and retail executives who took part in the Pulse Survey said their companies have been changed forever.1

Put demand-side data to work

Forecasting consumer demand has never been more challenging than it is in today’s environment. The Consumer & Retail sector, which has already seen a flattening effect on forecast accuracy metrics over the past decade, is now seeing additional accuracy erosion caused by the dramatic shifts in consumer behavior. Companies must challenge traditional time-series and regression forecasting methodologies and look to further develop and expand both their digital and analytical approach to predicting consumer demand.

Harnessing data from mobile, social, and e-commerce platforms and turning this data into insights that inform and enhance the demand picture are becoming table stakes. Furthermore, companies that use these internal and external data sets together with advanced statistical techniques and machine learning approaches can substantially improve current forecast accuracy, up to 50 percent.

Per the KPMG 2021 CEO Outlook Pulse Survey, a sizeable majority of leaders have accelerated new digital business models and revenue streams (69 percent).

One clear outcome from COVID-19 was the acceleration of online sales. In the middle of the pandemic during maximum lock-down, online purchases saw a step change reaching 19 precent of retail sales.2 KPMG projects that e-commerce as a share of retail sales will continue to experience upwards of 12 percent year-over-year growth.3

COVID-19 is an accelerant to online sales

E-commerce as a share of retail sales

Source: KPMG Economics, Census Bureau (September 2020), Haver Analytics
Buying behaviors of the digitally connected consumer are breaking the relationship with historical sales patterns and require a demand forecasting paradigm shift.
Ken Spigarelli, KPMG, Supply Chain Consumer & Industry Lead

Subscribe to our Consumer & Retail Insights program

Industry insights and trends, one click away

Industry insights and trends, one click away

KPMG believes companies that adapt their demand-planning processes to the changing landscape the best will be in a position to identify, respond and predict shifting consumer behavior. The shifting landscape requires consumer product and retail businesses to rethink traditional demand planning methodologies and to begin incorporating new techniques and capabilities into their demand cycles including: 

  1. More frequent analysis: Consumer online buying patterns are more erratic and increase demand variability. Companies that accelerate the demand review process will allow for faster recognition of shifting demand signals and help minimize impact on the supply chain.
  2. Planning hierarchies and attributes must expand: The e-commerce channel requires the inclusion of a new set of attributes into the demand forecasting process to enable better analysis and reporting. Companies using new hierarchical relationships and data points between product, customer, and channel will drive better correlation and segmentation along specific demand characteristics.
  3. Develop a data strategy for digital: Businesses should be purposeful when incorporating the e-commerce channel into their master and transactional data strategies. Companies must identify critical data elements and associated governance processes required to enhance the demand forecasting process, as well as those that are necessary to support effective reporting and analytics.
  4. Harness the power of advanced analytics: Let the data and power of advanced technologies work for you. Forecast accuracy has already stalled across the industry. The e-commerce channel has only added another layer of disruption. Companies that can leverage internal and external data sets, as well as a combination of advanced statistical techniques and machine learning approaches, will become more responsive and resilient. 

Client story

Bringing innovation to the table

Though it has been bringing iconic products to the global market for more than 125 years, this global food product manufacturer was not content to rest on its laurels. A KPMG proof of concept showed our client that it could leverage advanced analytics and predictive modeling to gain the insights needed to make better business decisions and adapt to changing consumer preferences.

The proof of concept enabled predictive demand modeling through the inclusion of two core elements. These being:

  1. Advanced modeling techniques that combined statistical models and machine learning.
  2. The incorporation of external signals, such as commodity price indices, weather cycles, and natural disasters to identify correlation to units sales, earnings and margin forecasts. 

The resulting forecast improved accuracy by 50 percent, cut 30 to 40 percent of the time required to create forecasts, and increased visibility into the drivers of their business.

AI and machine learning, combined with advanced modeling techniques, are expected to refine the company’s planning models and continue to improve their ability to learn and predict over time. Based on this success, we’re now helping our client expand its modeling across brands and varying operational and financial scenarios.

In our next blog …

We’ll dive deeper into how advanced analytics combined with machine learning capabilities can be used to enhance the demand forecasting process and fuel faster and better decisions.


  1. KPMG 2021 Consumer & Retail CEO Outlook Pulse Survey (January 2021)
  2. Source: KPMG Economics Census Bureau (September 2020), Haver Analytics
  3. Source: KPMG Economics Census Bureau (September 2020), Haver Analytics