Despite the transformative change sweeping the consumer goods industry, success in the analytics practice still relies on its oldest asset: people.
That was the somewhat ironic key takeaway from the 2019 edition of the Retail & Consumer Goods Analytics Conference, which kicked off last week in Chicago with a milestone: the use of a robot named Pepper (courtesy of HCL Technologies) to introduce the first speaker.
But although the planned agenda would have suggested that the adoption of artificial intelligence, machine learning, advanced analytics tools and other new-age technology would dominate the discussions, most of the presentations and subsequent conversations at least broached the subject of “human” personnel.
Based on the discussions, traditional consumer product manufacturers and retailers are working hard to identify the new skills required of a data-driven organization; to attract the new, often younger talent they need to fill those skill sets in a highly competitive job market; to re-educate the existing workforce not only on new technologies, but also on new strategic methods of problem solving; and to figure out how much of the current workload can be effectively “re-assigned” to machine learning and other automated tools.
One of the most critical of those new skills — storytelling, the need for analytics professionals to effectively translate data into digestible forms for business users and top management — was fully demonstrated on the stage through a procession of speakers who used tall tales, anecdotes, parables and the occasional off-color remark while offering their personal best practices in designing an intelligent, analytics-informed modern enterprise.
Among the more entertaining of the storytellers were a trio of presenters from Mars, Inc. Event Co-Chair Tarun Kataria, the company’s global director of advanced analytics and machine learning, told an ancient Arabic parable about three sons who are unable to properly divide 17 camels among them as their father had directed until adding, then later subtracting, an 18th camel. The moral of the story being that “analytics is the 18th camel” that can drive success for organizations. (Read the full story here.)
Meanwhile, Mars chief digital officer Sandeep Dadlani (below) modified an existing joke about a lost hot air balloonist asking directions from a pedestrian to illustrate the oft-troubling disconnect between the analytics team and the rest of the enterprise. (Read the joke here.) Replace “engineer” with “data scientist” and “management” with “CDO” to get Dadlani’s gist.)
Finally, Romain Apert, vice president-chief information officer for Mars Wrigley Confectionery, harked back to the glory days of TV game shows by posing the “Monty Hall problem” to illustrate the need for professionals to “learn to relearn, or even learn to unlearn” the analytic practices that have been guiding the industry until now. (See more below.)
Throughout the event, speakers and attendees alike acknowledged a similar shift in traditional thinking among both manufacturers and retailers that, ultimately, will lead to organizations fully guided in their go-to-market strategies by advanced analytics.
One significant example came from Walmart’s Enterprise chief information officer Clay Johnson, who noted that chief executive officer Doug McMillon has charged the company to give all employees the necessary understanding of new tools and technologies rather than using language they can better understand — because the time has come for everyone to become part of the intelligent enterprise.
“[We were] talking about cloud, artificial intelligence, machine learning — all these different buzzwords. One of the business leaders said, 'Hey, we need to figure out how to take those words and put them into terms the business can understand,’” explained Johnson. “And Doug stopped and said, 'No, we actually need to go the other way around. We need to force people to learn this terminology because it is what mainstream is.' That right there was a mind-shift. That set the tone, not only for the c-suite, but for the rest of the company.”
The following is an executive summary of the event’s full agenda:
Day 1 Workshop Unlocking the Future: Using AI-Driven Analytics to Determine Corporate Strategy
Derek Smith, Vice President of Services, Prevedere
Attendees at the opening day workshop agreed that while defining the business problem at their organizations can be relatively simple, integrating the right data and insights arguably takes the longest amount of time.
About 64% of respondents to Prevedere’s 2019 “Bridging the Data Divide” survey indicated they are not satisfied with their company’s ability to understand the impact of external factors on performance. One way to combat this is to transform existing analytics to be more-forward looking in a strategic roadmap.
Executives noted that, to see a full transition from a reactive to a predictive enterprise, companies need to change their internal culture. “The change management sometimes is harder than the system integration,” said Smith.
Day 1 Data & Analytics Share Group
Kapil Dabi, Partner/Principal for Digital, EY
Patrick Moriarty, Managing Director, EY
To succeed in the current environment, companies must focus on the human element to bridge the gap and move from insights to action to value. This was among the topics of the opening-day Data & Analytics Share Group hosted by EY. The critical success factors to drive positive returns from data & analytics are business need, flexibility, data privacy and user-centricity, explained Dabi.
How Walmart is Designing the Intelligent Enterprise
Clay Johnson, EVP & Enterprise CIO, Walmart
Andy Walter, former VP, IT & Shared Services, Procter & Gamble
During a “fireside chat,” Johnson explained how the retailer has created a Digital Transformation Office chaired by CEO Doug McMillon to make sure the c-suite is in the know and on board with IT and other related projects. Among the many endeavors supporting the enterprise, Walmart is testing the use of robots in 100 stores to scan for inventory (and even sweep the floor) while not being disruptive to shoppers.
As the retailer evolves its e-commerce and online grocery offerings, one challenge it faces is striking the right balance between the use of stores as shopping locations and fulfillment centers. Other topic areas Johnson discussed included Walmart’s use of predictive analytics/machine learning to hire better employees, and its growing partnership with Microsoft: After signing a cloud-deployment agreement in 2018, the partners opened co-located space for employees in an Austin, Texas, office that has a startup culture. “You can’t tell who is who,” he said.
Concerning competition with Amazon, Johnson said that Walmart has experienced success once it shifted from being defensive to becoming offensive.
Driving Digital Transformation at Mars
Sandeep Dadlani, Chief Digital Officer, Mars Inc.
Dadlani revealed a new purpose for the company: “The world we want tomorrow started with how we do business today.” Furthermore, “‘Digital Mars’ accelerates shaping the world by empowering associates to create value 100 times faster.” The Mars Digital Engine, Dadlani says, finds the problem, solves the problem and automates the solution. The company has 11,000-plus associates engaged in design thinking. “Once we have the right questions and reframe the problem, we use data and analytics to find the right answers,” he said.
License to Disrupt: Using a Digital Factory to Drive Change
Remco Brouwer, SVP, Digital Innovation & Strategy, Randstad Holding
Brouwer detailed how HR staffing/consulting firm Randstad created a “Digital Factory” in 2017. It was built next to existing structures in the company, reporting directly to the CEO. The multifunctional and multicultural team operates as a startup. Its mission is to scale the best ideas quickly around the enterprise. It serves as a model for a new way of working across the enterprise by delivering or failing fast, learning fast and capturing the learnings. Among the challenges Randstad has faced are the difficulties in driving change; accepting that assumptions from the past might not be valid today; learning how to drive scale and speed across a fragmented enterprise; and navigating the journey to become a truly client- and data-centric organization. Change management is the key to realizing benefits, he said.
A VC’s Perspective on the Potential of AI
Lonne Jaffe, Managing Director, Insight Venture Partners
Artificial intelligence has generated headlines everywhere – even theatrics – as companies try to capitalize on the hype. Computers are creating their own models of the world, and data is critical to the success of new models. But there are limitations and challenges that still exist with AI. Among them are the fact that data availability is still highly manual for many companies, it’s still difficult to trace or explain a decision, and privacy concerns limit AI applications. Investing in AI requires technical expertise to avoid common pitfalls and theatrics.
A Startup’s Perspective on Analytics
Brady Duncan, Co-Founder, MadTree Brewing
In creating a purpose-driven company, MadTree Brewing uses data analysis and word visualization techniques through an applied analytics traits assessment tool to attract ideal employees and grow. Duncan detailed how MadTree wants to hire the right people and understand them; refresh the company’s purpose, values and vision while rallying stakeholders around them; and monitor/measure progress through an equity health check of employees, partners and consumers.
Attacking the Biggest Business Opportunities with Advanced Analytics
Andy Walter, Former VP, IT & Shared Services, Procter & Gamble
Eric Chen, Director, Analytics & Data Science, Unilever Asia
Maria Macuare, VP, Data & Analytics, Campbell Soup Company
Mark von Oven, Former VP, Analytics, Target Corp.
Reiko Yoshida, Data Science, Facebook
In a discussion led by Walter, panelists tackled topics such as consumer and shopper engagement, engaging c-suite executives, measuring success, and talent development. Macuare said that CPG companies are doing a very mediocre job leveraging customer data, although they do well with their own consumer data. In dealing with executives and getting their buy-in, Chen said it’s important to understand their business strategies and plans, enroll them as fully committed sponsors, get moving on projects as quickly as possible, and get them implemented quickly to illustrate impact and value. Then, when it goes well, step back and let others talk about it.
Yoshida explained the ideal situation she has at Facebook, a bottom-up company where the individual teams doing the work are empowered and there’s no executive-driven approval process. Von Oven (who also served as the event's Retail Co-Chair) said it’s important to justify your existence and show how much value you provide to the organization. Where talent is concerned, Chen said, “We’re a scarce resource. Use that to your advantage. Engagement with senior leadership is critical.” Added Macuare: “Data scientists are the ones closest to the problem. Use them as a source of innovation.”
Supercharge Your Demand Planning & Forecasting with Machine Learning
Rick Davis, President, DDG, LLC (sponsored by SAS)
There’s a changing environment for demand planners, but one constant is that organizations need and expect improvements in forecast accuracy, Davis said. Companies need to identify opportunities to leverage machine learning to perform remedial tasks. IA – intelligent automation – will set new standards of quality, efficiency, speed and functionality that allow demand planners to be elevated into more strategic roles.
Building a Foundation for Intelligent Collaboration
Joe Wright, Lead for Integrated Business Planning, Kellogg North America’s Integrated Center of Excellence (sponsored by E2open)
Wright explained how the CPG manufacturer undertook an effort to improve the collaborative planning process with key retailer partners. After identifying the three main “pinch points” in its existing process — residing in the areas of product data, commercial analytics and supply chain analytics — Kellogg found ways to drive cross-functional analytics and cross-channel consumer insights (across category management, sales, marketing, supply chain, operations and executive management) to improve inventory management, reduce out of stocks and increase sales.
Key Takeaways from the 2019 Retail & Consumer Goods Analytics Study
Pete Reilly, SVP of Sales, Marketing & Services, AnswerRocket
Larry Layden, Client Partner, Capgemini
Mahesh Kumar, CEO & Founder, Tiger Analytics
Using the Retail & Consumer Goods Analytics Study as a backdrop, the panel discussed the efforts being undertaken by manufacturers and retailers to move from a product-driven to a knowledge-driven business focus. While progress is being made in the development of both internal analytics capabilities that can inform and direct the entire enterprise and mutually beneficial methods of data sharing between manufacturers and retailers, there is still a lot of work to be done, the panel concluded. But the ongoing implementation of AI tools should have a dramatic impact in both areas — provided the human side of the industry is ready for change.
Navigating the Marriage of Machine and Human Learning
Romain Apert, VP, CIO, Mars Wrigley Confectionery
Introduced by Pranay Agrawal, CEO, Fractal Analytics
The consumer goods industry’s traditional value equation model is failing because “it required that the world remain stable and predictable,” noted Apert. Today, companies must adopt a “deep and empathetic consumer centricity” to find the real problems that must be solved. But while machine learning will have a dramatic impact on the industry'stransformation, the “humans” involved also need to change the way they learn by shifting from deductive to inductive reasoning, finding the correct hypothesis within data rather than using data to prove (or disprove) the hypothesis. “We have a tendency to try and prove ourselves right” rather than letting the data guide us, he said. “We have not been educated to say, ‘I don’t know.’”
Unlocking the Power of Unused Data
Jamie Lancaster, VP Contact Center of Excellence, Kroger
With the rise of digital and omnichannel shopping, Kroger’s contact center has become rich with unused shopper data, or “micro data” as Lancaster calls the roughly 25,000 customer contacts it receives per day (via phone calls, emails and even snail-mail). It is useful data, nonetheless, that can be used to address specific local issues but also, in the aggregate, to educate and potentially influence Kroger’s store operations and other business activities. “What we lack in macro-level quantity of data, we make up for in micro-level quality of data,” he said. One hypothetical use: Being able to trigger product recalls much faster internally based on customer contacts rather than reacting to information from outside the company.
Improving Cross-Functional Collaboration Across the Supply Chain
Pradipta Saha, Associate Director, Supply Chain/Finance, HR Analytics, Mondelez International
Steve Sigrist, VP, Customer Services & Customer Supply Chain, Newell Brands
Among the industry practitioners who are less enthused by the idea of more data is Sigrist. “Anything that we do in supply chain needs to follow a basic formula, and the basic formula is Return = margin × velocity,” Sigrist said, adding that he spends much of his time focusing on the velocity side.
It’s critical for companies to build tools that can effectively reduce the number of issues that arise and quickly address the ones that do because “the best customer service is no customer service,” which implies that there are no problems to solve, Sigrist said. Advanced analytics can help smaller manufacturers that aren’t category leaders earn a seat at the table with retailers because “You’ve got to earn the right to solve those problems,” he said.
Keeping Pace with the Ever-Evolving Consumer
Fiona Swerdlow, VP, Research Director, Forrester
Though only about 14% of total U.S. sales are transacted directly online, more than half are influenced by digital in some way, according to Swerdlow, who noted that “the store is evolving and matters hugely, still.” However, digital now unlocks opportunities to directly target and influence individual shoppers wherever they are, and helps marketers (as well as Forrester) see consumers from a different vantage point: their behavior, rather than their demographics. The five key retail technology implementations that Forrester is tracking are: omnichannel, personalization, analytics, digitization and AI.