A brief history of the last decade in analytics: Big data is here! Now what do we do with it?
Insight isn’t an inevitable outcome of accumulating vast pools of data on customers, markets, or decision journeys. Many organizations struggle to turn analytics into a strategic capacity. According to a survey of executives by Forbes Insights, just “13% of companies can be considered ‘leaders’ in leveraging customer data.” The strategic research arm of the media firm goes on to warn, “Companies that are not adapting are finding their customer bases precipitously eroding as data-savvy competitors offer cheaper, faster products and more personalized experiences.”
Scott McDonald, CEO of the Advertising Research Foundation, writes in Forbes: “Businesses appear to be enthralled—but also intimidated.”
He cautions that the rush to data can easily go too far. Traditionally, insights departments have talked directly with customers through surveys, focus groups, and shop-alongs. “They have prided themselves on telling the inconvenient truth to misguided internal teams who advocate for a new product, a new ad campaign or a new strategy that turns off consumers.” Today passive sources of digital information can overwhelm or even replace actual customer interaction. McDonald isn’t a data skeptic, but he writes, “The best research and insight departments are heterodox in their search for data inputs, but judiciously evaluate the benefits and liabilities of those inputs and help their companies make wiser decisions.” That is, there’s a place for data, judgment, and perhaps even talking with consumers.
To understand how a larger consumer company is navigating these changes, Yale Insights talked with Stephan Gans, senior vice president and chief insights and analytics officer for PepsiCo. He cautioned that effectively using data for business ends requires a “blend of art and science.”
Q: Where does insights and analytics fit into the PepsiCo structure?
Historically, the role of insights and analytics in PepsiCo, as in most companies, has been a marketing function. Often, it has been a support role doing evaluation and testing. Today, businesses and brands need foresight. The insights function has to evolve to provide a forward-looking strategic advantage for the company. That’s the transformation that we’re driving at PepsiCo.
Q: How does that happen?
It’s not easy. Change needs to happen in insights and in the adjacent functions, like marketing, R&D, and innovation, as well.
Within the insights function, we’ve got people, processes, tools, and the organization. We’re driving change along each of those axes. We’re changing the number of people and the type of skills that we allocate to certain roles, particularly moving to more specialization over jack-of-all-trades roles. We’re locating insights people directly in the brand teams. We’re simplifying our processes and digitizing all the key tools that we use to enable standardization, globalization, and better leverage of scale. We’re making the organization a lot flatter.
Insight lets you do things better, fast, and cheaper. It lets companies make progress on all three fronts simultaneously, especially in a big company like PepsiCo.
Q: Is PepsiCo on the vanguard or is this happening across industries?
There’s certainly a lot happening in insight. There’s brilliance at every aspect of insights somewhere—very specialized companies that offer completely state-of-the-art niche services—but it’s very rare that you come across brilliance at everything within one company.
I intend for PepsiCo to really make a difference in the way evaluative research is done today. Because of the scale that we have, we can have a real impact. I think in the next two years, we’re going to see some really big shifts in this industry.
Q: Can you give us an example of an improvement that has grown out of the changes in customer insights and analytics?
Here’s a small example. Insights let us understand what’s happening in the split second when people make a decision about what beverage to pick. We’ve found that in order to meet our performance with purpose targets, that is, to inspire people to buy products with less added sugars, sodium, and saturated fat, a small change, just rearranging the products in a vending machine in such a way that they are a more logical fit with people’s state of mind and the way we make choices, goes a long way. We can get a high double-digit impact in the percentage of people that choose low-sugar or alternative beverages, which helps achieve our targets in creating a more nutritious portfolio of products.
Q: Do insights and analytics ever lead you down the wrong path? When do judgments come into play?
Great question. In terms of testing new products or whole new brands, we apply very sophisticated tools that take all sorts of scenarios into account, but in reality, all the parameters we plugged in change between the test and launching the product or the brand. Maybe the product itself is slightly different. Or the distribution.
That means the expectation of the success curve of a product will be too high or too low. You’ve got to be careful about taking those tests too literally. I find blending data with judgment leads to the best results.
Q: Are there areas where, no matter how much data you have, you can’t reach 100% certainty?
If you look at creative material—commercials, advertising, or pieces of content—it can be very surprising to see what goes viral and what doesn’t. Two out of three times, the things that become a huge global hit are the last things I would have expected to become so omnipresent and so successful, while the things I expect to take off don’t necessarily get the same pickup.
It is so hard to predict. Again, I am a strong believer in the blend of art and science—judgment and data analytics—ultimately producing the best results in terms of forecasting as well as understanding.
Q: What’s the role of the insights and analytics function in PepsiCo’s shift to more nutritious products?
We’re very closely involved. It’s such an ambitious pivot that all of our brands are involved. Again, this is an example of a situation in which you need to blend qualitative judgment that’s rooted in experience with the story the data tell us.
In a research setting, if you ask people, “Are you interested in a more nutritious version of a product?” they will be very positive about it. But in practice, people will resort to the option that they’re used to consuming and buying more often than you would expect, based on their input. So again you’ve got to make sure that you’re aware of the bias.
Q: How do you manage customer insights in a global company that operates in so many different contexts?
The fundamental operating model of PepsiCo is to leverage our global scale while keeping our businesses very focused on relevancy in local markets. There are universal human truths; however, those universals are very high level and abstract. For a brand to market its products effectively, you often need to adjust the way in which you bring that general consumer insight to life to a local culture.
I like to talk in terms of building global might for the global fight. But much of the competition is local; the negotiations with channels, retailers, and even media are often local. The challenges are local, but our business is global. So how can we leverage our scale globally while leaving room for local relevance?
It requires a careful balancing act between those things you can actually leverage globally—a general positioning for a brand like Lays, for example—while the execution against that brand positioning really needs to be different in, say, Pakistan versus Argentina. That means you need the capabilities in the central brand group to make sure that that global positioning is as universally relevant as possible without it becoming vanilla. And you need capabilities in Argentina and Pakistan to really execute against that global positioning in the best possible way for that local market. It’s something we are constantly readjusting for.