CPGmatters: Welch's Focuses on Shoppers to Increase Category, Brand Sales

Let's start by being skeptical about 29% category growth, especially attributable to new buyers. That would mean that (a) squeeze bottles are a huge segment of the category and (b) all of a sudden people noticed they were there--people who weren't buying this category in these stores before.

That said, the syndicated suppliers have been using this data this way ever since the widespread availability of scanner data, comparing chains in a market on brand and product development. This is not a new insight or a particularly new tool, although IRI may have made the navigation of the data simpler. We used to have a different tool--it was called client service.

Dr. Stephen Needel, Managing Partner, Advanced Simulations

The results of the analysis by IRI and Welch's had a grain of luck in it. Welch's was not producing the variety (of package type) that some consumers were looking for. It seems to be a clear-cut case of if you don't have what consumers want, they'll go somewhere else.

But uncovering the missing product to increase sales is not always the solution. The client may have all the SKUs. And so the manufacturer must be ready to dig deeper into consumer behavior, not just in terms of transactions but how the decisions are made in the store. It's a matter of better understanding how the consumer thinks about the category before going into the store and then observing and examining (eye tracking) shopper behavior in the store. It's another approach to shopper insights.

This is grunt work for some in that you have to roll up your sleeves and be willing to go along with the consumer to grasp their feelings about she experiences to understand how it influences their decisions in that retail store, i.e. to buy products there or go elsewhere.

Joan Treistman, President, The Treistman Group LLC

To truly understand your unrealized sales opportunities, you need data at both the macro and the micro level. At a macro level, market data and analysis from IRI, Nielsen, or any number of consulting firms can help you understand the level of sales you should expect from a particular category given the specific attributes of your chain. And, as described in this example from Welch's, if there's an obvious assortment problem, this can help you find and fix it. The far bigger opportunity, however, can be found at the micro level--specifically, in the frequent shopper data.

Today, virtually everyone shops multiple retailers, so there will be category voids for most of your shoppers, including many of your best customers. If you can find the shoppers whose overall purchases suggest they should be buying from category X, but whose actual purchases in that category are less than expected, then you can selectively market to those shoppers to try to broaden their baskets. Or, you can use the same analysis to find a set of shoppers to survey about the reasons they don't buy certain products from your store.

Although this type of targeted discounting and surveying can pay huge dividends, the traditional barrier has been the difficulty of doing the analysis. Most tools for behaviorally identifying and segmenting shoppers are cumbersome, expensive, and slow; and because they can only be used by a few highly-trained analysts, the turn-around time for this type of work is often days or weeks.

We believe that as a new generation of technology enables faster, easier analysis to be performed by anyone in the organization, we will start to see an explosion in highly-targeted communications from retailers and manufacturers, which will result in dramatically higher profit per shopper.

Ben Sprecher, Founder and VP, Marketing, Incentive Targeting, Inc.

I agree with Mr. Needel - a 29% increase? How? Did they add to the total number of SKUs within the category? If so, how many? Did they increase the overall space devoted to these products and thereby change the customers' perception of the category within the retailer's stores? How much of the increase was due to the addition of the squeeze bottle SKUs? The implication is it is from new customers entering the store. If so, what happened to total sales?

Scan data from an individual chain tells you what customers bought, but not what they might have purchased had the items been carried. Comparing data from different retailers in a market does allow you to gain insight on what they purchased elsewhere, but should not be interpreted to mean that they would have necessarily purchased those items in the same quantities from you. Customer profiles and preferences vary and while this article focuses on specific items it does nothing to address why a customer picked one retailer over another.

Steve Montgomery, President, b2b Solutions, LLC

This type of growth may seem incredible, if not impossible to replicate. My research shows that when retailers put the consumer at the center of every decision, they will realize similar results. I've led studies that examine SKU duplication and the possibilities associated with elimination of duplication...for example, the typical category can shed 15% of the SKUs (as in actual SKU count) and realize improved sales and profitability. Image what you can do if you take it to the next level with assortment optimization by store.

True category localization will deliver sales and profitability that retailers have yet to experience. There is a challenge with this concept. Just exactly is meant when you hear "localized assortments"? To what degree of granularity is associated with a localized assortment? How are the associated processes built in a way that delivers localization but does not require an extreme amount of increased workload on the merchandising staff? These are the questions I am researching and uncovering.

It is clear that there is an area that retailers must be careful to avoid. If you understand your customers through your own data, you are only looking at a portion of the available market. It is critical to examine the entire set of consumers that are within that radius of a reasonable drive time to a given store. This is where syndicated data and other sources must be included to avoid a myopic view of consumer demand. When retailers combine the two with true localized assortments, pricing, promotions, and space allocations, the results will be nothing short of phenomenal. There is no grain of luck here...its about identifying what consumers want to buy, not selling products to consumers.

Kevin Sterneckert, Research Director, Retail, AMR Research

This type of consumer insight analysis is fine, but it misses the critical question: what can I do that truly changes what the consumer buys? Too often, the analysis ends at the point of identifying a fact (affluent shoppers buy more squeeze bottle jams and this particular retailer is targeting that segment). That's a great start, but so what? What are you going to do about it? I could think of a hundred ideas. Which of those hundred ideas work, for Welch's and the retailer? Which just shift demand within the category with no benefit to the retailer? Which shift demand between retailers with no real benefit to Welch's? Which just push margin dollar into the consumer's pocket without any incremental spend? These are the tough questions.

Jonathan Marek, Senior Vice President, Applied Predictive Technologies

Welch's doesn't welsh on facts. But certain facts can have more value to some implementers than others. If the Shopper Insight Explorer process is as effective as projected above, and I have no reason to refute their projections, it will probably become widely used. But if all major CPG companies use it and all major retailers participate, will a win-win sales panacea happen for everyone with sales increasing to new levels? Or will some other condition arise?

Gene Hoffman, President/CEO, Corporate Strategies International

I have a friend in retail who keeps saying, "what's wrong with retail today is that we've got too many scientists involved." And I agree with him. The IRI research sounds pretty strong, but the caution would be that it replaces actually getting the Welch's brand managers out there THEMSELVES to listen to and watch customers and associates. There is no substitute for that experience.

Eduardo Castro-Wright, Walmart exec., has a nice interview in the New York Times Business Section this past Sunday, where he talks about the best strategies being derived from talking to customers yourself. I agree with him in that at times, we all try to make retail much harder than it really is. I hope Welch's is thinking like a customer and not like a scientist. After all, we are all customers, just put that hat on!

Lee Peterson, EVP Creative Services, WD Partners

Today's marketplace has become extremely complex, with consumer products companies struggling to gain insight into consumer and customer needs, and to execute their business practices with precision and agility. Increasingly, companies are turning to data analytics as a tool to drive deeper, "right-time" insights, enabling them to quickly and intelligently respond to changing market demands. However, the way many companies are conducting analytics and developing insights today is often inconsistent in quality or is limited to only certain parts of the organization. This may be the case here with Welch's.

The issues companies face begin at the very source with data. With an explosion of available data in the marketplace, consumer products companies are faced with an array of data formats and new data integration and management requirements. However, even as the amount of data available expands, consumer products companies continue to struggle with gaining access to retailer data, which provides a level of granularity not found elsewhere.

Historically, consumer products companies have relied primarily on syndicated data sources, providing data in standardized formats, to understand general purchasing trends. However, today, these companies need to develop a more granular understanding of shopper behavior and market dynamics.

Luckily, the amount of data available to companies has increased tremendously, including enterprise data, radio frequency identification (RFID) data, web data, unstructured data (e.g., e-mails, text documents), and in many cases, POS or loyalty card data directly from the retailer. Together, these sources represent the key to unlocking a better understanding of shopper behavior and market shifts. The issue, however, is that these sources exist in a wide variety of formats and are difficult, if not impossible, to analyze together. To effectively integrate a variety of internal and external data sources, consumer products companies should both align enterprise data and promote the development and adoption of data standards. From an internal perspective, enterprise data management capabilities provide the people, processes and infrastructure that drive internal data consistency, process integration and the accuracy of business information within the enterprise. Whenever an original source of data is changed, enterprise data management capabilities ensure that the change is replicated across the organization.

Shopper and Category insights has been an emerging field for the last few years and has not yet been mastered by most of the CPG Manufacturers. Some manufacturers have set up Shopper Marketing Groups, complete with a Director and as many as 15-20 analysts, that are tasked with generating shopper and category insights. Their customer is various groups that sit in either the Sales organization or in Marketing, and sometimes both! They include Category managers, Sales Executives, Merchandising Managers, Researchers tasked with New Product Development, Pricers, Channel Management and Trade Promotion Managers.

Bottom line, there are too many tools for analysis, and still too many decisions being based upon "gut feel". Take a step back. Analyze your data. Go visit your stores. Have progressive collaborative workshops with the retailer and CPGer at the table. Balance the science with the art of category management. There is no silver bullet. Not yet, at least.

Ralph Jacobson, Global Consumer Products Industry Marketing Executive, IBM

I agree with my colleagues that the increase seems implausible, but my question would be, how does this increase in assortment affect related categories such as peanut butter, crackers, etc? Does this change increase overall basket size? Does it increase customer satisfaction? What about the purchase cycle? If the squeezable bottle lasts longer than the jars are they really increasing share of wallet? If the price point is higher than jars are they not purchasing other products in the store to compensate? These are the real questions retailers care about.

Michael L. Howatt, Retail and Shopper Insights Consultant, Ascension Growth & Innovation Strategies

If Shopper Insight Explorer had never been invented, would anyone at Welch's have noticed that there were no squeeze packages on the grocer's shelves?

Mark Lilien, Consultant, Retail Technology Group