PepsiCo's Data Analytics for Business Growth

PepsiCo, a global consumer packaged goods giant, faced significant challenges in the 2010s to maintain its historical growth pace. To overcome these hurdles and unlock new opportunities, the company embarked on a transformative journey leveraging advanced data analytics. This article provides a comprehensive overview of PepsiCo's Data Analytics for Business Growth, detailing how it revolutionized their operations and customer relationships.

Unlocking Growth: PepsiCo's Data Analytics Strategy Explained

In 2019, PepsiCo was a $65 billion global company with 267,000 employees, generating substantial revenue from its North American divisions like PepsiCo Beverages and Frito-Lay. Despite its vast reach and strong brands, achieving non-dilutive growth became increasingly complex due to market changes like the explosion of private labels, new consumer preferences for fresh brands, and diverse sales channels. Michael Lindsey, Frito-Lay Chief Transformation and Strategy Officer, highlighted the challenge of continuing growth in a non-dilutive way.

PepsiCo's traditional growth pillars included:

• Manufacturing a large variety of well-known products (22 brands generating over $1 billion annually).
• Achieving broad reach with products in hundreds of thousands of locations, supported by the largest truck fleet in the US.
• Offering a strong value proposition through economies of scale and low costs.
• Creating an appealing market presence with promotional campaigns.

Facing a comparatively low-growth North American market, PepsiCo leaders sought to use its scale and product variety as a competitive advantage. Their new goal was to identify latent consumer demand and niche opportunities through data and advanced analytics, pinpointing where to place specific SKUs and how to communicate with interested shoppers.

The PepsiCo Demand Accelerator (DX): A Centralized Analytics Powerhouse

To drive its data-driven transformation, PepsiCo established the PepsiCo Demand Accelerator (DX) in early 2015. This centralized unit was tasked with creating and delivering standardized, enterprise-wide data-driven marketing services. Reporting to the Chief Commercial Officer for North America, the DX brought together approximately 180 people from various units like Category Management and Shopper Marketing.

Jeff Swearingen, Senior Vice President of the PepsiCo Demand Accelerator, explained the need for speed and connection: "We looked across the marketplace and recognized that we needed to move quickly. We realized we had pockets of strength in data analytics, but they weren't well connected. We needed to bring those together, democratize them, scale them, and get more leverage from them."

Initially, the DX was allocated startup funds, with the expectation to become self-funding over time by generating incremental revenue and capturing cost synergies. Swearingen set three main goals for the DX:

1. Drive change in internal brand and commercial strategies.
2. Elevate PepsiCo's position with retailers in terms of partnership and data analytics reputation.
3. Motivate PepsiCo retail partners to make better business decisions using new shopper insights.

DX Organizational Structure and Capabilities

The DX was structured with distinct units performing complementary activities. In 2019, it comprised five units:

• Shopper Marketing: Created programs for national promotions or specific retailer needs.
• Shopper Analytics and Insights: Observed shopper behavior and leveraged advanced analytics.
• Category Leadership: Collaborated with stakeholders to drive product demand and shopper loyalty.
• Space Optimization and Transformation: Developed store-level assortment and space recommendations.
• Media and Data (added in 2019): Developed capabilities for personalized shopper connection.

Early success stories included the launch of Quaker Overnight Oats, where DX data and analytics identified 24 million US households for engagement, leading to 80% of its sales growth in the first 12 weeks. The DX also prioritized solving important, solvable problems that unlocked sales growth and gave PepsiCo a competitive advantage, ensuring retailers would act on the insights.

Advanced Analytics Platform and Shopper Data Insights

The DX's technical needs were supported by two major IT efforts. First, PepsiCo Information Management had been developing enterprise data efforts since 2012, establishing a data analytics taxonomy for a "single version of the truth" across 37 countries. This led to a product hierarchy, master data management, and an enterprise data warehouse.

As demand for advanced analytics grew, IT leaders realized the need for a new data lake to complement the data warehouse. Martha Roos, VP, IT, Digital Services Delivery, noted, "We were purposefully slow to embark on a data lake. We didn't want it to become a data swamp. We made sure we had a strong use case that really needed this data lake."

Second, PepsiCo Enterprise Architecture collaborated with the DX to build an enhanced digital services capability, an "anchor project" creating momentum for new investments. Together, they established the Next-Generation Advanced Analytics Platform (NGAA). This cloud-based platform could source and manage larger, more diverse data from internal and external sources, providing secure access for retail partners. The NGAA was rolled out incrementally within six months.

Most Valuable Shopper Data and Machine Learning

The NGAA provided access to "Most Valuable Shopper" data, a core DX activity. This 20-terabyte database contained 106 million anonymized shopper records with over 1,800 attributes per record. Data was amassed from US census data, anonymized shopper-level profiles, weather data, and product shipments. Households were scored with product probability scores (e.g., likelihood to buy a category or brand).

Greg Lyons, Chief Marketing Officer, PepsiCo Beverages North America, emphasized understanding people: "We needed to understand people, inspire people, and value people. And if you look at our data journey, a lot of it has been trying to understand people—what makes them tick, what’s important to them, what are their consumption habits, etc."

This granular data allowed DX associates to identify latent shopper demand, differentiating even between stores on opposite sides of the street. Initially, DX used business intelligence reporting and statistical modeling. Under Michael Cleavinger, Senior Director of Data Science and Advanced Analytics, they moved to machine learning for sophisticated predictive models.

One example involved maximizing fountain drink sales for a convenience store and gas (C&G) retailer. Since scan data only reflected cup purchases, DX combined syrup usage data (PepsiCo's and other providers') and developed ensemble machine learning models to identify store and shopper attributes influencing syrup usage. This resulted in ranked lists of syrup recommendations for different C&G store profiles, implemented in over 1,000 stores.

PEPWORX℠: A Suite of Data Analytics Solutions

By 2019, PepsiCo packaged ten of its solutions into PEPWORX℠, a suite of data analytics capabilities. PEPWORX℠ was offered as a complimentary service to help PepsiCo and its retail partners personalize product assortments, optimize store space, manage marketing dollars, and launch innovative marketing programs. It helped differentiate PepsiCo in the marketplace by unlocking granular sales growth.

Key Use Cases and Benefits of PEPWORX℠

PEPWORX℠ helped drive various benefits for partners, including faster product turns, better profit realization, and higher net price realization due to less need for discounting. Shoppers benefited from tailored access to valued products.

Examples of PEPWORX℠ in action:

• Brand Accelerator (Cheetos Asteroids): Using social media data and shopper profiles, the Media and Data team identified convenience stores most likely to serve the niche audience requesting the discontinued product. This precise approach led to "far better than expected" sales.
• Space and Optimization (Planograms): DX associates developed an approach to tag individual stores with identifiers reflecting their local shopper base. They created planograms (diagrams for product placement) for about seven distinct store profiles per retailer, advising on total category allocation, not just PepsiCo products. Ankur Meghani, Senior Director, Space Optimization and Transformation, noted the success: "I’m proud of how we leverage insights like store-level clustering—and how those insights have become the language of the business."

Transforming Retailer Relationships: From Transactional to Collaborative

The DX capabilities became central to PepsiCo's Joint Business Planning (JBP) process, an annual formal approach with largest retailers to mutually drive growth. Bryan Santee, Vice President, National Sales, PepsiCo Beverages North America, described the shift: "We use a clock as a metaphor for where you are with your customer in the relationship. 1:00 to 3:00 on that clock is a very transactional relationship... As you move around the clock, the relationship becomes more collaborative... With the help of the DX team, we’ve shifted with many of these customers much farther into the collaborative side of the clock."

PepsiCo's success measures evolved to reflect a "three-audience win": simultaneously delivering for the shopper, the retailer, and PepsiCo. This approach significantly influenced PepsiCo's reputation as a partner. Kantar's PoweRanking, which surveys retailers, recognized PepsiCo as the #1 partner in North America for four straight years, attributing it to "better solutions from us now that we’re acting like one PepsiCo instead of many fragmented divisions."

Custom Solutions and Turnkey Applications

The DX often designed custom solutions for large retail partners with complex problems. This involved assessing the retailer's "will and skill" to implement changes. They used proofs of concept and A/B testing to validate effectiveness, even gaining access to some retailer data to tailor solutions. For instance, they helped a retailer optimize marketing programs for thousands of stores, considering unique local needs.

When custom solutions proved scalable, they were converted into turnkey applications, often with user interfaces. The Quaker Overnight Oats project, for example, initially took three weeks and four people to generate a list of "Most Valuable Stores." After automation, this was reduced to three days, and ultimately, a user interface delivered the capability in thirty minutes. These turnkey solutions also allowed the DX to serve smaller, regional retail players more efficiently.

Continued Evolution and Future Potential of PepsiCo's Data Analytics

By 2019, as the DX reached its fifth anniversary, leaders encouraged a "Year One" mentality to foster continued innovation. They aimed to expand the influence of shopper insights, integrating them more proactively into PepsiCo's end-to-end supply chain. This included leveraging Customer Supply Chain capabilities like the Demand Signal Repository (DSR) and cloud-based tools such as:

• FAST: A mobile frontline sales reporting and alerting application for sales associates.
• Innovation Tracker: A dashboard for tracking SKU distribution and performance.
• PromoX: A software application to monitor national or regional promotions at the store level.
• Test and Learn Dashboard: Used to monitor and measure the outcomes of various treatments.

John S. Philips, Senior Vice President, PepsiCo Customer Supply Chain and Go-To-Market, stated that these efforts created a connected supply chain with customers, enabling partnerships to drive store-level execution and sales. Kirk Tanner, CEO, PepsiCo Beverage North America, highlighted the DX's future potential, especially for the beverage business, aiming to be "nationally great, locally even better" through greater customization and agility at the local level. The DX's impact was expected to grow by connecting more directly with regional divisions to tackle local opportunities, strengthening retail relationships and overall growth.

The DX also actively educated PepsiCo associates through a DX portal with learning materials, successful use cases, and online training, averaging 10,000–15,000 file downloads per month. DX University further provided training on various team capabilities. The model's success led to its replication in European and Latin American sectors, operating informally as a global unit to share best practices.

FAQ: PepsiCo's Data Analytics for Students

What is the PepsiCo Demand Accelerator (DX)?

The PepsiCo Demand Accelerator (DX) is a centralized unit established by PepsiCo in 2015 to develop and deliver innovative data-driven marketing services across the enterprise. Its main goals are to drive change in internal strategies, elevate PepsiCo's reputation with retailers, and motivate partners to make better business decisions based on shopper insights.

How does PEPWORX℠ contribute to PepsiCo's growth?

PEPWORX℠ is a suite of data analytics capabilities launched by PepsiCo in 2019. It helps PepsiCo and its retail partners personalize product assortments, optimize store space, manage marketing funds effectively, and launch innovative marketing programs. By providing granular insights into shopper demand, PEPWORX℠ enables targeted strategies that lead to increased sales and stronger retail partnerships.

What kind of data does PepsiCo use for its analytics?

PepsiCo leverages a vast array of data, including anonymized shopper-level profiles, US census data, weather data, and product shipment information. This is combined into a "Most Valuable Shopper" database with over 1,800 attributes per record. They also use granular daily scan data from retailers, managed through systems like the Demand Signal Repository.

How did data analytics transform PepsiCo's relationships with retailers?

Data analytics, particularly through the DX and PEPWORX℠, transformed PepsiCo's relationships with retailers from transactional to collaborative partnerships. By offering data-driven solutions that helped retailers optimize their businesses and improve profitability, PepsiCo earned trust and became a more consultative partner, recognized for helping achieve a "three-audience win" benefiting shoppers, retailers, and PepsiCo itself.

What role does machine learning play in PepsiCo's data analytics?

Machine learning is a key technology and methodology used by the DX to create sophisticated predictive models based on its extensive data. It allows them to tackle complex retail and category problems, such as optimizing fountain drink sales by identifying specific store and shopper attributes influencing syrup usage, leading to highly granular and effective recommendations.