Adjust Font Size :

Pundit’s Pulse Of The Industry — Park City Group’s Randy Fields

Debbi Fields is well known for starting up the Mrs. Fields cookie shops. The success of that first shop is interesting, but for many the more interesting question is how did she grow and manage the operation so that it would successfully encompass almost 500 shops in the U.S., Hong Kong and Singapore.

Part of the answer is having the right husband. With a background in business and technology, he sought to find a way to grow the chain while maintaining the quality standards that his wife implemented in that very first store which she personally managed.

Of course, this dilemma of how to manage multi-unit operations while maintaining quality standards is a high priority for fast food chains and, indeed, any business that seeks to duplicate itself in many locations.

Not surprisingly, the confluence of data and technology are often seen as crucial tools to manage far flung operations.

Now we’ve run many pieces focusing on category management. In addition to an interview with Bruce Axtman from the Perishables Group, which you can read here, we have run interviews with both Willard Bishop Consulting’s Bill Bishop, Information Resources’ Thom Blischok and a joint interview with FreshLook’s Mark Degner and IRI’s June Fenzel.

The world is also filled with tools to help retailers and their vendors operate more efficiently and effectively. A particular product that is designed to help this process along is Fresh Market Manager, which is described this way:

Fresh Market Manager is a fully integrated, end to end, management system for perishable, non-perishable and food service departments such as bakery, deli, produce, meat, seafood, dairy, floral and frozen. FMM provides category analysis and precise item management, in-store forecasting and production planning, computer-aided ordering, perpetual on hand inventory and real time alerts.

This system is developed by a company headed up by none other than Mrs. Fields’ husband, Randy Fields. We asked Pundit Investigator and Special Projects Editor Mira Slott to find out more:

Randy Fields
Park City Group
Park City Utah


Q: You helped build a cookie empire. How did a cookie guy get involved in this business? Take us back to the genesis.

A: The defining moment that led to where we are today is when Debbi [Randy’s wife and founder of Mrs. Fields Cookies] said she was unwilling to open a second store. She was satisfied with the success of the first store. My suggestion came from the perspective of a business guy. It was obvious we could build a significant chain and business. Let’s open a thousand of these stores. I asked her, why just one? ‘I can’t be in every store, so I’ll never be sure employees are doing it the way I’d like it done,’ she said. ‘Will they care as much about the quality? Will they know how much to make? How do we control it?’ Debbi wanted Debbi in every store.

With my background in artificial intelligence, computer science, and decision theory, I knew there was a technological solution. The idea went through my mind that we could build technology to embody the same “business sense” as our founder. That system could be deployed allowing our best manager to be in every store. I used to call it a human Xerox machine; in essence take your best managers, find out what they did to succeed and build that in the system, which would be employable across different units. That was the concept.

We’ve been on that path ever since, frankly a remarkable venture to say the least.

Interestingly, the very same idea that grew from Debbi’s first store has been carried forward to today. In every business there really is an expert whether operator, manager, merchant, who knows the best way to do what has to be done… how many bananas to have out in the perishables department, etc. Inevitably it comes down to someone extremely good at what he does. We build those rules into the technology, in essence creating computational control of those activities that can be applied in every store.

Q: So there are Debbi’s entrepreneurial business skills and technological, mathematical formulations for determining the amount of cookies to make, by variety and for different parts of the country, and during holiday times, maintaining consistent recipe and quality and freshness throughout the chain, etc.

A few questions come to mind:

Point One: In the cookie business compared to the produce industry, you’ve got relatively limited number of SKUs, finite variables to control, certain ingredients that will be the same in all products in all locations. A produce department deals in hundreds of SKUs and categories, a wide range of commodities, value-added products, and multiple supply sources, all influenced by Mother Nature, seasonality, short shelf lives, and a host of other factors. A freeze hits the Florida orange groves or a buyer gets a good deal on a shipment of tomatoes, etc., and all bets may be off.

A: We hear these concerns often. Whether bakery or produce, we use the same technology. We had a chain of Mrs. Fields bakeries with hundreds of SKUs. But the same idea applies regardless of the number of SKUs. Variability of supply is important. Let me qualify. It adds ordering complexity, such as which supplier do you take? We had multiple suppliers for ingredients, different flours to procure, and seasonal differences affected the cookie. Mother Nature impacts the cookie business too. Cocoa goes through the roof because of a freeze. You find a way to cope with the problem.

What is it our customers are coming for; is it quality or simply availability? Let’s say there is a major problem with tomatoes. Would you sell terrible quality tomatoes? No. While these issues create aggravation, they don’t change the end result. In point of fact, many businesses don’t have a strategy laid out, which can lead to confusion. If you’re the tomato buyer, are you required to keep tomatoes in stock or just quality tomatoes?

The mechanics of the supply chain change but not the end result. Plannograms change, pricing changes, but not strategy. You are already dealing with tremendous fluctuations store to store. Chocolate chip cookies with milk chocolate were difficult to sell on the east coast, but the west coast was a milk chocolate market. Brownie sales were significant in one store while in another store it was muffins. We took those differences into account and built them into the technology. The technology enables us to manage differences, not sameness. If a supermarket has a 100 produce departments it tries to plannogram by clusters, but truth is that variability is much greater than that. There is a much greater degree of micro marketing to specific areas.

Q: POINT TWO: On one hand, you have a savvy produce manager who has honed the art of selling product based on his experience and instinct, perhaps an in-depth understanding of his customer base, etc., and on the other, you have complex data and historical analyses beyond the limited perspective of one person on the ground. Does knowing how many bananas to put out, at what ripeness, accommodating corporate pricing strategies, and promotions rest more on the latter?

A: You’re pointing out the complexity of business. The merchant making one set of decisions, operators implementing tactical decisions at store level, the problem sets. Category management often doesn’t have enough information that’s actionable or the deep analytics to make decisions from his vantage point. Another problem set occurs at store level, the tactical decisions they have to make are often very computational, administratively intense and prone to error. In between this, there is the failure to synchronize what the category manager decided and what store and department managers are implementing. We don’t know any supermarket that can tell you other than visually that managers put out what they were supposed to.

Animals perceive uniqueness, humans perceive sameness. The human brain is built for pattern recognition and tends to consolidate to one model. The power of the computer is individual differentiation; the ability to look at 110 stores as if they’re unique. One store might need four trays of milk chocolate and another might only need one.

Q: POINT THREE: So there are regional and market differences, variances on a store-to-store basis you’d need to take into account. At the same time, Debbi’s original concern was that consistency in quality and concept could be extrapolated and maintained during a major chain expansion.

A: To build a brand you cannot have uniqueness in every store. There has to be commonality and then some degree of micro marketing. If you went to different McDonald’s and no two products were the same, it wouldn’t work. Take the bakery or produce department; 80 percent may be determined by central merchandising and 20 percent by the local manager, or reverse, even if highly decentralized.

Q: Produce branding is not as common as in other industries. In the same way, retailers face challenges in creating a unique produce department brand. Could you elaborate on your concept of brand identity?

A: Here’s an example: If a retailer’s deal is 39-cent bananas, it doesn’t matter if the price of bananas is different in other markets. The brand stands for 39-cent bananas. If they don’t sell, then you throw them away. Retailers need to build brands within the retail umbrella. What does your produce department stand for? You need technology to support local, individual implementation of the brand. Someone has to establish what the brand stands for.

The next granularity is technical — how many do we need so they remain fresh, but not so many that shrink occurs or pitch is not sustainable. It’s mathematical and shouldn’t be delegated to the person in charge of customer service, etc. The role of technology is to take the strategic vision of the merchant and insure the strategic assortment is being maintained and take over the tactical challenges of what quantity at the right time.

Q: In category management, item-level sales results relate to and are impacted by overall category sales, by adjacent commodities, seasonality issues, other categories throughout the store, such as snack items in the dry grocery aisles, etc. At the same time, a specialty product that might not sell large volumes could be important to the store’s best customers, or most profitable customers. Does your system adjust for these scenarios, and if so, how?

A: The system customizes based on what you’re suggesting. There are different sets of criteria. It can be optimized by profitability, volume of items, consumer demand. The manager could say they want to select top line items, or bottom line items.

Those issues you bring up are true but not as important as people imagine. Most critical influences within a category are availability and the promotional plan. If in fact the consumer trades chocolate croissants for chocolate donuts, it is measurable and forecastable. The technology picks up the promotional impact or availability impact of a product. If you promote raspberries and it causes apple consumption to fail, we can measure that. People may fall back on general statistical theory and say there are 82 factors that influence raspberry sales. Even if there are 82 influences, a few variances will account for 70 or 80 percent of the change. Even if it affects sales of yellow mustard, it’s not important.

Say raspberries in season reduce apple consumption. It doesn’t impact things so much. Our system could take all these different factors into account, but there are only a top few influences you need to track and monitor.

Q: Take us through the steps.

A: First, we start with very granular store data, scan data, what’s selling, what’s not, what pitched, what ordered, what profitability, item-by-item, store-by-store, day-by-day. We put all this information in the hands of our analysts and in the hands of the category manager to help him adjust assortment, promotional planning, etc. We help influence results by doing the deep dive analytics.

Second, at store level, we use the same data, and literally are able to make hour-by-hour decisions on what product should be available, the quantities, all ordering, production, etc.

Third is synchronization. We link these two pieces of technology so the store is implementing the plan of the category manager. There’s a feedback mechanism so operational hierarchy can figure out what happened. We’ve addressed the whole problem in this way.

Q: How do you get all this information?

A: Our systems connect with all the systems store by store, and work from there. We get it from the merchant, or the operational systems, financial systems connect. This is next day information, store-by-store, scan-based from the retailer. Some information we’re getting from the suppliers. When you have retailers and suppliers working together, out-of-stocks go down.

Q: Do you analyze this internal retail data in context of competitors, regional markets, etc.?

A: We don’t see much value in using outside market data. Our focus is on the data that drives the retailer’s bottom line. We help retailers and suppliers understand the dynamics of category profitability from a store-by-store, item-by-item perspective. It would be difficult to effectively manage against market share unless you fully understand what is going on within your own categories. We uncover the details for “fixing the mix” and determining how to make the kind of profits needed to run the business.

Q: So comparing your performance in different produce categories to that of your competitors is not important?

A: In the really big picture, market share analysis doesn’t really prove helpful. General Motors used to care about market share and kept selling cars at a loss and losing money. You could say our competitor has a larger mind share of produce. If 63 percent of people out there think the supermarket has better produce, that’s where information could help. Mind share versus market share; very different concepts. The most important factor isn’t the market share data. It’s your own economic performance. Market share is not a call to action.

We can demonstrate that no one who has used our tools has failed to make more money. Our customers in every case have become more profitable; the reason for that is relatively simple: what we do is deep dive analytics into data that the retailer does not routinely review, such as scan data, costs, etc. We convert data into money, that’s our business mantra, find golden nuggets in the data to actually help companies make more money. Our value-add is that most merchants do not have the time (and often the tools) to do as much analysis of their business as they would like to do. We do that for them, and then reduce all of the analysis into action plans to improve profits.

Q: Could you share case studies you’ve done related to produce, deli or seafood to give our readers a more in-depth understanding? Have you conducted cost/benefit analyses on the impact?

A: Our relationships with clients are proprietary. But we have some case studies you could share with your readers. (Editor’s Note: The case studies can be accessed here.)

Q: Could we go back to your thoughts on the value of outside data?

A: Market share data is an idea worthy of discussion. Suppliers are by and large interested in share if they are dealing in commodities and there is zero difference between their product and that of their competitors. There is no value-added in a commodity business. If that is the case, to increase share, they buy more and sell cheaper. It hurts the retailer. There is a certain inherent structural problem. When suppliers are commodity-share oriented, it pushes retailers to be commodity-share oriented. All the suppliers have figured out is that commodity business is bad. Value-added product is needed.

But even more importantly, collaboration between the supplier and retailer is essential for each to become more profitable. Retailers need suppliers to help them improve their competiveness and suppliers need to help retailers survive the “value retailer” onslaught.

Q: How do you view your role in this dynamic?

A: We see our role as mediating the trading relationship between retailers and suppliers, bringing a third-party independence to help answer questions like: what should the assortment plans be, what should the promotional strategies be, how much should we be ordering to reduce out-of-stocks and to avoid excessive shrink.

The supplier wants to improve volume and reduce out-of-stocks. The retailer is interested in that, and in reducing shrink. The retailer has limited amount of space so he wants the highest level of productivity per square inch or foot. The supplier wants as many SKUs in full distribution as he can get.

They need an independent third party to mediate what SKUs are in what stores. Maybe star fruit sells wonderfully in 30 stores. Why should it be in 100 stores creating shrink? The supplier will be in trouble with retailers. We help by determining what’s profitable where and at what times to rationalize SKUs and cluster SKUs appropriately.

Sometimes high velocity items are out of stock because low velocity items are causing shrink. There’s been such a proliferation of bagged salads, sometimes the highest velocity items are out of stock, while some exotic blend doesn’t sell and goes to shrink. What we do is analyze down to the SKU level continuously day-to-day, as well as historical data, so we know what sells in a particular area.

Q: How do you translate strategic analysis to tactical execution on the retail floor? Are you involved with store-by-store implementation and monitoring of these programs?

A: The Park City Group takes a phase approach. We want to begin by working with category management; look at one to two years of historical data and do deep dive analytics, what works what doesn’t. For a few months, that strategy is articulated and what happens from a merchandising and operational strategy is consistent.

Then we move into the store with synchronization. We install systems in store, look at item-to-item data, and do hour-by-hour forecasting, inventory movement, shrink management, ordering, placement, etc.

The last piece is helping people with labor schedules. To have the right product in the right quantity and the right time, employees need to be doing the right thing at the right time. We facilitate the process, having the capability to help category management with tools and analysis, going into stores to work with production. We are an enabler of vendor managed inventory.

Q: Wal-Mart was known as an innovator in vendor-managed inventory. How does it compare to other retail models and with what you do? Are you finding more interest and demand from retailers for vendor-managed-inventory schemes?

A: Wal-Mart provides immense amount of store-level information, scan and inventory data so vendors can see how their products are doing across the Wal-Mart system. They are responsible for keeping Wal-Mart stores appropriately in-stock. Most vendors have teams of people looking at data, how much product sent where and how much is needed to be in stock.

We do something similar, but can also provide forecasting and ordering on behalf of the supplier and retailer. The retailer then brings suppliers into the loop with us, and is ultimately responsible for paying our compensation. There’s a trend in the industry toward this; were seeing more interest in vendor-managed inventory, and our job is to facilitate and make this easier.

When a company uses our technology and service, it doesn’t matter who is responsible. Most suppliers have no idea how to build a forecasting model for managing inventory at a retailer. Both need better ordering systems and category management systems to be more profitable together.

The difference with Wal-Mart is that Wal-Mart makes data available to the supplier and says you need to be responsible for how much product has to go through the Wal-Mart system to get to store 56. We can help the supplier generate the order. Our systems are sophisticated to improve the ordering process and reduce shrink. Wal-Mart isn’t concerned with how the supplier gets it done.

We can work with retailers and suppliers on the issue of promotional strategy. They know how many cases moved but was the promotion profitable? Often the promotion is not well executed and analyzed. Banana sales go through the roof and the manager orders the same amount as last week and it results in major shrink. The ad comes out Wednesday with strong sales and the store runs out of stock by the weekend. The manager scrambles, gets a big order and is stuck with cases of bananas because the promotion may have been overly successful. Through our analysis, we know that on week 27, the retailer runs bananas 29 cents per pound and we know how much will move.

Q: Have there been any surprises in the work you’ve done, any unexpected results? What is the most unusual information you’ve discovered?

A: A lot of people focus on the shrink problem. The biggest surprise with everyone we’ve worked with is that if shrink is high, out of stocks are high too. What they don’t realize is that if the person doing the ordering doesn’t have the right information, they are making mistakes at both ends. Better forecasting is needed.

Q: Could you discuss what’s happening in produce departments more specifically?

A: Produce and perishables have become incredibly important in terms of differentiating retailers. Because it’s become so much more important, there’s a need for more information, better strategy and execution. How are you going to be more profitable? As center-store margins contract and are under pressure from big box retailers, produce has to become more profitable. Let’s simplify the job.

Take a basic business problem: The chain does fresh-cut fruit in store and wants to grow the category. The produce director wants the fruit in certain configurations and containers. How do you communicate that to the stores? Presently, the director may send out an e-mail or memo.

When using our technology, the category manager made this decision, goes to our technology and delineates this group of stores do watermelon in quarts and pints and honeydew-cantaloupe combinations this way, etc. We coordinate and synchronize what happens at the store level and the category manager receives feedback. The next day, the category manager knows how many quarts and pints were made and sold, what time to restock, etc. Now he gets real-time answers and feedback to what occurs on the retail floor.

We come from a retailer background. A category manager in retail doesn’t have the time to do deep dive analytics. We come to their aid, not just by providing tools but analytics. The problem with operators is they don’t have the right information and are focused on 50 things. When all pieces work together in the big scope, it’s stunning.

The money quote in this interview is clearly this:

We can demonstrate that no one who has used our tools has failed to make more money. Our customers in every case have become more profitable; the reason for that is relatively simple: what we do is deep dive analytics into data that the retailer does not routinely review, such as scan data, costs, etc. We convert data into money, that’s our business mantra, find golden nuggets in the data to actually help companies make more money. Our value-add is that most merchants do not have the time (and often the tools) to do as much analysis of their business as they would like to do. We do that for them, and then reduce all of the analysis into action plans to improve profits.

We strongly suspect it is true and Randy is honest enough to point out that it is not the program that does the magic but, rather, the “…deep dive analytics into data that the retailer does not routinely review…”

In other words, if you run your business based on a hunch or what your boss taught you, the business won’t do as well as if you really understand the business.

Besides, in most cases there is simply no alternative any more. It is not as if one option is to have highly experienced and knowledgeable produce managers and the other to use technology. Many operators today don’t pay enough to hire high level people as produce managers and then retain them for decades as they gain experience. For many operations, it is technology or nothing.

At the same time the technology is, practically speaking, more limited than we might hope. Stores today are so large, carry so many items and the response to many initiatives so dependent on promotion and so delayed as consumers absorb information as to the availability and use of products that few indeed are the retailers who really build an information system to know what to do in all circumstances.

Debbi Fields wanted to make sure that cookie-makers in her franchise stores would know when to bake and when not to — that is not easy to determine.

Yet produce, and the perishable departments in general, pose special difficulties because these departments are expected to do more than maximize sales and profits — they are also the front page for most stores, the differentiating factor.

So when Randy says, “They need an independent third party to mediate what SKUs are in what stores. Maybe star fruit sells wonderfully in 30 stores. Why should it be in 100 stores creating shrink?” — we say maybe.

Toys R Us was built on selling Pampers as a loss leader to bring in customers with children crying for toys — perhaps some star fruit in the produce aisle, like garnish on the plate at a restaurant, has an effect on consumer attitudes toward a store that is difficult to quantify.

Of course, none of this means data shouldn’t be used; it just means it shouldn’t be used crudely.

Certain aspects of this program strike us as major advances over often-used practices. When Randy talks about labor schedules, we want to say hallelujah: “The last piece is helping people with labor schedules. To have the right product in the right quantity and the right time, employees need to be doing the right thing at the right time. We facilitate the process, having the capability to help category management with tools and analysis, going into stores to work with production. We are an enabler of vendor-managed inventory.”

We have seen all too many operations handling this backwards — avoiding a strawberry ad, for example, because labor is unavailable, rather than having the flexibility to analyze whether the profitability of the promotion will allow for the labor to be scheduled.

Of course, now that so much product is sold packaged — such as strawberries in clamshells — avoiding a strawberry ad, a trend only certain to increase with country-of-origin labeling, we suspect that the drive will be to treat produce departments more like dry grocery, which will mean less willingness to invest in produce managers with special expertise. This means we will need analytic systems such as Fresh Market Manager even more.

Whether for retailers looking to run their operations more efficiently and profitably or for vendors looking to handle replenishment in a way that retailers will increasingly demand, these types of systems are clearly the wave of the future.

We appreciate Randy Fields and the Park City Group for taking the time to discuss this important technology.

Print Friendly, PDF & Email

The Latest from Jim Prevor's Perishable Pundit