Kraft and Gatorade Examples Deliver Actionable Results
Insight to Action’s depth of experience with customer segmentation consulting led us to distill these five success principles, with three tips from the Kraft cheese division and two from Pepsi’s Gatorade and Propel brands.
Kraft Customer Segmentation Consulting Example
Kraft President Bob Eckert chose The Cambridge Group as its customer segmentation consulting partner. Under Eckert’s leadership, and that of the Kraft insights lead Sue Ellen Bohac, this work was exceptionally well adopted and drove brand plans.
The goal of the customer segmentation consulting was to provide marketing focus for Kraft’s cheese brands, as a critical input to the annual brand plan development.
The Cambridge Group consulting team worked with over 10 different Kraft brand teams, including Philadelphia, Velveeta, Singles and Cheez Whiz. We also worked with the Kraft Dips business, which was a smaller, non-focus business in the same organizational division as the cheese brands.
These three success principles emerged, which are helpful for any customer segmentation.
1. Encourage adoption with top down and bottom-up reinforcement
One of the tactics that Eckert used as a leader was to ask each brand team to frame their recommendations around the brand’s target customer segments. So, for instance, if a brand team recommended a specific consumer promotion, he would ask which customer segment this promotion addressed and how the promotion addressed the segment’s interests.
His consistent inquiry reinforced the segmentation work from the “top down.” If the brand team’s recommendations did not address the segments, the plan needed to be reworked.
Meanwhile, The Cambridge Group team worked “bottoms up” with each brand group to identify its customer segment target and to align brand actions with the target.
The Kraft team observed that repetition is important for adoption in a large organization such as Kraft.
One Kraft executive elaborated that a success marker for a study or initiative could be measured by the number of copies requested by other leaders. At that time, paper copies were still common, and The Cambridge Group’s silver binders full of customer segmentation materials were frequently found in the offices of Kraft’s brand managers.
When the segmentation was initially introduced to the division in a large meeting, my colleague Gloria Cox and I needed a furniture dolly to wheel in the 100+ required paper copies to distribute.
2. Invest in large sample sizes and test multiple segmentation approaches
With a nationally-representative sample of over 3,000 consumers, this customer segmentation consulting effort required multiple segmentation models, given the breadth of the cheese category and Kraft’s portfolio of natural and processed cheese.
For example, cheese is used as:
- a cooking ingredient
- a snack on its own or with other foods
- a topping or sandwich enhancement
While small by “big data” standards, the 3,000+ ending sample meant that study data could be “sliced and diced” into smaller groups with confidence.
The broadest customer segmentation framework was the cooking and cheese segmentation. An example customer segment that emerged was the Involved Cheese Cooks who both enjoy cooking and like cooking with cheese.
This inclusive segmentation and customer target (i.e., Involved Cheese Cooks) was relevant for a number of brand teams who use cheese as a cooking ingredient, such as Velveeta, Cheez Whiz and Kraft natural products like Kraft Shredded cheese.
On the other end of the spectrum was the cream-cheese-specific segmentation, which was tailored for the Philadelphia brand group. This was a distinct segmentation that could be cross-referenced with the cooking and cheese segmentation. If memory serves, there was a Cream Cheese Lovers segment that was the primary consumer target for the Philadelphia brand.
One of the major advantages of the large sample sizes that the 3,000+ ending study delivered was providing robust ending samples for each customer segment. For example, with six segments who fall roughly evenly in the population, each segment will have an ending sample of 500.
When brand teams wanted to look at customers within a segment who are, for instance, heavy users of the brand compared with medium, light or non-users, the subsegment sizes could be counted on to be sizable. Too often, samples sizes are kept as minimal as possible to save cost (for instance at 1,000 or 1,500), which means that this type of subsegment analysis won’t be very reliable.
Another example of the benefits of large ending sample sizes was a segmentation for AT&T that included lookalike modeling to type its prospect database.
3. Integrate customer segmentation with purchase behavior data and broader data sources
Adding to the Kraft study’s depth and complexity was a decision made on the part of the insight team to tie the segmentation work into consumer diary work from National Eating Trends (NET) data.
This supplemental data came in the form of printed binders, and since this study had over 10 banners, this amounted to over 100 binders. I recall the bookcases that needed to be put in place to house the collection.
This diary data allowed us to look how a segment recorded use of cheese in-home for a two-week period. I remember being somewhat bemused at the time to learn that the ham and cheese sandwich was still a number-one ranking item for lunch for many consumers.
Actual purchase data was calibrated from Kraft’s panel providers, either IRI or Nielsen. The customer segmentation market share was nearly identical between claimed and actual purchase data. We also learned that the stated purchase quantities for the consumer packaged goods (CPG) cheese products were over-stated by a factor of three.
We used a tool called Quanvert, which made it easier to interrogate the data for custom analysis that was not specified in advance in the tab and banner plan. At that time, Quanvert didn’t live up to the initial expected usage of providing Powerpoint-ready graphics, but it was helpful to slice and dice the data.
There was little interest on the Kraft team’s part to learn how to use Quanvert or how to dive into the diary data. The solution proposed was for the summer intern at Kraft to learn these techniques, a highly impermanent solution for knowledge transfer.
Within the carefully-negotiated scope of work that The Cambridge Group provided, my colleagues and I delivered the additional analytic cuts needed from Quanvert and NET data.
In 1994, roughly two years after the study was completed, the Insights lead, Sue Ellen Bohac left Kraft to create her own customer segmentation consulting firm. Despite the value delivered by The Cambridge Group, CPG clients were interested in a less expensive solution for customer segmentation, and her firm received many referrals.
Gatorade Customer Segmentation Consulting Example
Roughly 12 years later, our firm, Insight to Action was selected by Pepsi’s Gatorade and Propel brands for a customer segmentation consulting project to create a segmentation focused on active consumers. Approximately 60% of Americans at that time qualified as active, and this intentional sample allowed the team to focus on the most relevant consumers.
“to meet the needs of those who want to maintain a healthy lifestyle and want a lightly-flavored hydration option with vitamins and no artificial colors.”
The primary expected application of this customer segmentation was for innovation and new product development for both Gatorade and Propel. The lead Pepsi clients were Karyn Peterson and Sue Vanderwiel.
This customer segmentation example built on existing insights that Pepsi for “Athletes and Committed Exercisers.” And this example provides two more success principles for customer segmentation consulting.
4. Map demand with both customer segment and need state perspective
As is common with beverage segmentation, both need states and consumer segments are important to leverage. For beverages, the specific drinking occasion or need state is critical. As we previously explored, Gatorade built the brand on replenishing the athlete’s body during and after exercising.
In this case, need states were defined as “why people use the product in the moment.” Each consumer will have several consumption occasions or usage need states during a week, a month or even the same day. For instance, an athlete will typically have more occasions related to active thirst than a desk-based worker. The definition includes how, when, and whether the customer chooses to use the product. Both functional and emotional drivers are integrated.
A simplified need state example comes from a recent NPD convenience store need states study. NPD found used four major benefit areas, i.e., fueling (31%), wellness (23%), connecting (22%) and gratifying (24%) that led to specific need states.
NPD’s convenience store fueling need states included:
- “easy on-the-go” and
- “don’t feel like cooking”
And NPD’s gratifying need state examples were:
- “comfort foods” and
- “reliable favorites AM”
5. Focus on the relevant occasions
Meals are important, and so it might seem that the right focus for a beverage would be around accompanying a meal. However, in the case of Gatorade and Propel, we made the decision to focus on non-meal occasions. We also excluded occasions where enjoyment was the primary focus.
The team chose the following definition for relevancy:
“Recently consumed a food or beverage in support of your performance, energy, health or wellness goals – excluding foods or beverages consumed purely for enjoyment.”
Using this focused lens, six “macro” need states were identified:
- sports/exercise performance improvement
- weight management
- boost energy
- basic quench (hydration)
- preventative nutrition
- treating the body right
In addition to the six “macro” need states, the innovation team utilized 19 “micro” need states from this customer segmentation consulting work.
One need state where the Gatorade brand has strength is sports/exercise performance improvement. A complete profile was developed including: the benefits sought, the situation, the solutions (in addition to Gatorade) and the customer segments.
These five success principles from our experience and seven outputs with impact, are intended to provide some starting guidance for brands that are seriously considering a customer segmentation initiative. Additional value-added techniques abound, such as The Cambridge Group’s palate mapping approach.