CPGmatters: Coke Focuses on Segmentation For New Go-to-Market Strategy
By John Karolefski
Through a special arrangement, what follows is an excerpt of a current article from CPGmatters, a monthly e-zine, presented here for discussion.
The right product in the right place with the right message for the right shopper. That’s the goal of Coca-Cola’s new go-to-market strategy that focuses on a segmentation platform to address shopper needs and deliver customized beverage solutions to supermarkets.
“We start with the premise that no two stores in America are exactly alike. Each store has its own DNA consisting of different shoppers who have different needs and different purchase behavior,” John Carroll, Coke’s vice president of shopper insights and marketing solutions, said recently at Nielsen’s Consumer 360 conference. “Once you understand the store’s DNA, you will be able to recommend the right brands, right packages, and right occasions for the right people.”
In the past, the same products were largely available in every supermarket. Today, trading partners are looking for differentiation and segmentation, according to Mr. Carroll.
Coca-Cola’s Shopper Segmentation Methodology starts with shopper insights by using the Nielsen Homescan Panel of households and grouping shoppers into segments based on their purchase dynamics, demographics, and psychographics. That information is used to group stores into five clusters using Spectra targeting tools that identify high-value consumers. Coke scores each store according to beverage purchase segments. Once the stores are assigned a score, they are clustered together based on degree of similarity within the clusters and differences between them.
Each store typically serves one shopper cluster. Gaps in actual volume versus demand indices are opportunities. The latter estimates potential, indicating high consumer fit for the brand.
One challenge is taking Coke’s segmentation model and matching it up with a retailer’s own segmentation model. Mr. Carroll also admits that the system requires art as well as science.
“We’ve done our best to develop the science and to be accurate with clustering. But when the account manager walks into that Winn-Dixie in Cluster #1 in Florida or that Safeway in Cluster #2 in Northern California, they see that the clustering that we have identified as matching up with the people shopping the store. It’s really important to have those two things working together. So, we identify the shoppers in the store using eyeballs and the science and then we figure out what the merchandising differences are by cluster.
“That may mean loading up on Coke Zero in a suburban affluent store and Fanta in a Hispanic store. The process aims to figure out what that shopper’s needs are and what brands they are most likely to buy. This information helps Coke develop planograms for the shelf set, displays and cooler.”
Discussion Question: What do you think of Coca Cola’s new shopper segmentation strategy? What do you think are some of the opportunities as well as challenges in implementing such a system?
This was a great idea even in the early ’90s, when we built the tools to do this at Nielsen–this is not a new idea, even for Nielsen. Companies that do store-door deliveries have a unique opportunity to implement customized selections. There should be a good upside by improving the assortment.
The downside is that this is likely to produce less of a benefit than perhaps expected, because the world does not have five clusters and a store’s trading area population is never going to belong to one cluster alone. There will be some missed opportunities by focusing on demos alone and such a high level of aggregation.
The notion of taking a more customer centric approach is right but I agree with the previous comments–there seems to be a disconnect in the problem Coke is trying to solve. Carbonated beverages as a category is down and unless it’s down because stores aren’t in stock with the particular brand of soda the customer wants, I don’t see where this strategy solves that problem. Certainly delivering a more relevant tailored product offering may spur some impulse purchase but I don’t know if it’s enough to really move the needle and justify the added distribution expense of such tailored planograms.
The other issue raised here: whether all of one store can be identified in one cluster-was the first thing that struck me when I was reading it. The whole process starts with a very consumer centric approach but then seems to break down a bit when you say that every store has only one cluster–now your segmenting the store not the shopper. The segments sound suspiciously like the segments Wal-Mart was after before they backed off that strategy. I believe Wal-Mart realized the cost to execute was going to be too high. I wonder what Coke will find.
As the major beverage manufacturers fight for “share of thirst” it only makes sense to refine their assortment strategies to take advantage of the consumer purchase opportunities in the various retail and food away from home outlets.
The continual expansion of beverage options owned and/or distributed by Coke and Pepsico makes this a logical opportunity to maximize return on shelf/cooler space and specialized logistics for both the manufacturer and the retailer.
This in my opinion is a logical extension of category management and channel-specific packaging options that have long existed in the business…the major difference is it plays well beyond carbonated beverages and into water, juices, iced tea, iced coffee and other unique/upscale beverage choices.
I’m sure there is also new research to support this effort and boost sales on these important product lines!
Coke is a sage soda company. Its new shopper segmentation strategy places local product preferences in front of local consumers. And it deserves kudos. That approach is a logical market segmentation strategy and it will improve availability of locally desired products, increase the retailer’s shelf productivity and enhance product movement. That appears to address a lot of Coke’s marketing needs, but does it address another large company need–creating new customers’ desires for its currently new and its still-to-developed products for its future prosperity?
Coca-Cola should be commended for their impressive efforts to bring a more scientific method to the store segmentation issue.
In a recent survey conducted by Dechert-Hampe, Coca-Cola was widely recognized for its in-store marketing leadership. However, that same survey revealed that most marketers feel the barrier to more effective shopper and store segmentation was not just the insights, but the difficulty of in-store execution. Its extensive DSD network gives Coca-Cola a significant advantage in this regard.
We have learned that we can often develop more sophisticated models for shopper insights and store segmentation than the retailer can implement. So does that mean, we all need to go DSD? No, it simply means we have to use the shopper insights to continuously improve category management within the context of the systems available to us.
The basic idea has been around for a long time. Identifying the needs, wants, purchase behavior, and trends of consumers at a particular store and providing the products they want to purchase is the premise many successful independent retailers, some large retailers, and some major manufacturers are using successfully.
However, as soon as Coca-Cola says that this translates into determining the DNA of the store and that stores should be clustered into groups of 5, they have just negated the above premise and won’t find the success they desire.
Coke is really onto something here. Location-based marketing, or micro-geographic marketing, is smart business. Why? Because it capitalizes on the huge “under-the-radar” opportunity presented by the mass domestic migration happening in this country.
This past month, The Brookings Institute published a book called Boomburbs. Boomburbs dimensionalizes the scale of change in American communities. Here is an excerpt:
“Boomburbs are defined as having 100,000 residents or more…Boomburbs now contain over a quarter of all residents of small to midsize cities.
Another way to grasp just how big boomburbs have become, is by comparing their current populations with those of some better-known traditional cities. Mesa, Arizona is bigger than such traditional large cities as Minneapolis, Miami, and St. Louis…”
To understand who is living where–in center city, first ring suburbs, exurbs, boomburbs–is to understand your market. This is the lesson Coke is teaching.
When I think about all the planning that goes into DSD assortment rationalization I always think of a conversation I had with a route man one day. “A route man’s favorite is a route that ends with a couple of chain supermarkets.” Those are where you get rid of all the dregs that remain on the truck.
It’s great that our industry seems to be singing the right product/place/shopper/mantra and Nikki’s point is dead-on that the consumer is sure to benefit in the long run. In my mind, the more critical issue is that one that she calls out regarding the need for the CPG and retailers to better work together to figure out how to make this goal real. And, how to not only look in the “rear view mirror” but to utilize sophisticated behavioral profiling techniques to predict future purchase propensity as well.
What Coke is doing is not new…many CPG’s and retailers have been trying to crack this nut for quite some time. As an industry, we just care more at this point in time because evolving our store strategies is becoming a critical necessity rather than a “nice-to-do.”
As in-store media technologies evolve, it seems obvious that they will be ideal tools to deliver these micro-targeted communications. It won’t be easy to bring the appropriate vendors, internal constituents, processes and tools together to make this happen. But, it will happen in the near future and I applaud any retailer with this goal in mind.
I feel that there is room for the Coke way of segmenting and differentiating product selection by market.
For example, the A&P family of banner stores sells a huge quantity of both Coke and Pepsi products. They seem to alternate the WOW promotion weekly between brands.
For the July 4th promotion, marketing at A&P got creative and offered the following: if the customer purchased 2 12 pack cartons of Coke, they got 3 free. The retail price was $4.99 ea, so the customer got 5 12 pack cartons for $9.98. This stuff flew off the shelf. Coke and Diet Coke were still the best sellers, but all the others, including Fanta, Minute Maid and Sprite, did very well too. Even Safeway couldn’t beat the price. They had to feature it at 5 for $10.98.
It is a no brainer now with both Coke and Pepsi, and their access to updated demographics. They can tailor the mix and selection of their respective soft drink products down to the individual store.
But remember, the manager of a 65,000 sq. ft. or more super store has far more room for displaying product piled 8 cartons high on four or more pallets than the manager who manages a 30,000 sq. ft. or less store. That manager has to be creative, and having tweaked some of the mix, would go a long way to satisfy his/her customer base. Preferences, even of soft drinks, tend to matter.
Having too many Diet Sprite cases and no Diet Coke cases can signal that the store isn’t properly ordering the right mix of flavors and brands.
Having worked for KO for a number of year and still being a stockholder I would be shocked to see a single SKU eliminated from retail distribution. Does anyone think that Coke will eliminate Fanta Orange in any location that sells Nehi Orange? This is marketing–smoke and mirrors! Shelf space allocation rationalization might be a better term as I suspect that Coke will use this “segmentation” to try and convince retailers to eliminate all Orange drinks from a location so they can use the shelf space to expand their other labels. Hope it works because we all know shelf space = sales! If it were me, I would use the time and energy in building Coke in China where they have a large global brand lead and will get an even bigger push via the Olympics.
John Carroll, Coke’s vice president of shopper insights and marketing solutions, discusses the premise that no two stores in America are exactly alike and that the DNA of no two shoppers are the same. Whereas, I definitely concur that no two shoppers are the same, I respectfully submit that the correct premise should be, “…no two stores in America SHOULD BE exactly alike,” however, retail chains for the most part have become copy cats, followers, and dittoheads. The current retail universe is boring for consumers because too many stores are almost exactly alike. Most retail chains have 99% the same product assortment, the same types of promotions and TPRs, the same types of advertising, and in some cases, even the same signs, colors, and banners. I like the direction Coke is going to help bring back some individualism in how we market to different consumers. It’s a step in the right direction.
The entire industry is looking at better understanding shoppers, segmentation, local marketing, etc. and any manufacturer that is making strides in that area deserves kudos. The more targeted and closer to meeting the stated (and unstated) shopper needs and expectations one can be with planograms, pricing, communication, etc., the more sales will be generated and the higher the sense of loyalty.
What I see as missing here (the Holy Grail) is how to get out in front of the needs and not rely on previous purchase, neighborhood demographics, and other “rear-view mirror” approaches to predicting the future. While important to track and be mindful of, none of us drive cars solely by looking at what is happening behind us any more than we expect the coroner to assist us in our healthcare decisions.
The focus on shoppers is right on. The need to recognize differences between and among stores is correct. The segmentation idea is on target. It is what is feeding those distinctions that is “the best we have” currently, but not sufficient. I have been working with clients on improving upon this very issue and rather than place a commercial for my offerings here, I invite requests to discuss offline.
“The right product in the right place with the right message for the right shopper.” That’s a succinct definition of the holy grail of marketing and advertising, and targeting, or segmentation, is the way to achieve it.
Savvy, forward-looking retailers are adopting shopper-segmentation strategies specific to each of their stores. Some of these models center on customer demographics (who they are); others focus more on attitudinal attributes (what they want). The more a retailer can segment its own consumer base, the more actionable intelligence that will be revealed.
Strategies should focus on identifying opportunities, customer-targeted tactics, and benefit quantification.
Smells like customer centricity to me. It’s easier for Coke to pull off its own segmentations because of it’s a DSD product–they have more say over the product mix at a store by store level. What’s interesting though is that they have done this on their own, and now are working to reconcile their own segmentations against retailers’ segmentations. It’s too bad that retailers continue to be tight-fisted about shopper segmentations and insights–a manufacturer, with the broad view across all retailers, can bring valuable information to a retailer’s segmentation analysis, just as a retailer can help a manufacturer understand their unique customer mix–and potentially get a better product mix as a result.
But no matter what, with everyone paying such close attention to “the customer,” at least one group will clearly benefit: consumers.
Coke has been disappointed in their market share growth (or lack of growth) for years. Every 1% is very meaningful to them, since the carbonated beverage market isn’t growing. Pepsi bought noncarbonated brands (Tropicana, Gatorade, etc.), The noncarbonated segment has been growing nicely for everyone, not just Pepsi. If Coke’s new market segmentation strategy gets them a percentage point or more growth, then it’s a winner. The big growth would more likely be found by brand acquisition or substantial product innovation, however.
One of the key issues in activation of segmentation strategies today is the level of collaboration between manufacturers and retailers. Marketers such as Coke, who invest significantly in making segmentation actionable at retail, should be commended for their efforts. What’s needed, however, to really please the shoppers, is a true collaboration between retailers and manufacturers so there is visibility between the two–how do the consumer segments from the manufacturer match up with the shopper segmentation model from the retailer?
If the segments match up along a number of key parameters, the impact at store level can be significant. I hope that Coke can lead the way to help open the collaboration doors.
As a shopper, I am always delighted when I walk into a store area and it has obviously merchandised to match the kind of shoppers it typically gets. Grocers like Piggly Wiggly do a great job with this in their resort area stores–they are set up for the vacation stock-up trip and they are fun to shop. They put the pool and beach toys right up front with the soda displays! The portable cooler is right there too. Coke can help them sell more if they work together to understand crossover of segments to determine how much of each specific product they need to maximize the volume!