FEATURE1 October 2011

When it comes to the crunch


Jeffrey Henning of Affinnova on how techniques inspired by the theory of evolution helped Snack Factory extend its product range.


The challenge
In 2004 Snack Factory put a twist on the traditional pretzel by flattening it to create the Pretzel Crisp. The success of this healthier alternative to potato chips (some flavours contain no fat at all) helped make Snack Factory one of the fastest-growing companies in the US in 2009, according to Inc Magazine.

To build on its momentum, Snack Factory reinvigorated the brand last year, introducing new flavours and brighter packaging. This spring the team wanted to extend its range with two new ‘bold and spicy’ product lines. The brainstorming process produced hundreds of possible flavour and packaging variations that the marketing team wanted to explore.

To break into the crowded snack market, Snack Factory had created a speciality pack of Pretzel Crisps for airlines, while down on the ground it avoided the crowded snack aisle and sold in the deli department of grocery stores, where its novelty made it a popular impulse buy. Snack Factory knew that its new products would need to have packaging that supported this impulse buying. With so many variations of its ideas for new products, it was unsure which combination of flavour and packaging decisions would perform best. Manually winnowing the variations before launch would have introduced the risk of omitting the ones that would have been most successful.

The brief
Snack Factory needed to choose those concepts out of its 182 possible product ideas that would have the best shot at winning business from competitors like Baked Lays and Stacy’s Pita Chips – and do it in time to introduce the product at the International Dairy-Deli-Bakery Association (IDDBA) conference, just eight weeks away.

“This unique approach to research helps to put metrics behind the emotional task of determining which new products to offer. We have greater visibility into how they will perfom at retail”

Perry Abbenante, Snack Factory

The approach
Snack Factory feared that conventional market research methods would force them to arbitrarily restrict the number and diversity of concepts that they could afford to test with consumers. While the firm’s chefs had already developed recipes for many potential flavours, Snack Factory executives were most concerned with how the presentation of these new flavours would affect sales.

Because of this need, and the short timeframes, the company decided to turn to Affinnova, whose software applies genetic algorithms to the marketing process. This enables researchers to determine the best concepts from a large number. The first respondent chooses from randomly generated concepts, composed from attributes provided by the client. But later respondents react to concepts that evolved according to the choices of earlier respondents. This is based on algorithms that favour past respondent preferences while also incorporating random mutations. This survival of the fittest can determine the best concepts by interviewing just a few hundred online respondents.

For this project Affinnova surveyed two separate groups of 400 consumers to select the two product concepts most likely to succeed out of the 182 alternatives. Target consumers were in the 25-to-54 age range, had purchased snacks in the past two weeks and had said they were open to buying pretzels and ‘bold and spicy’ flavours. Key subgroups included current Pretzel Crisps buyers, healthy snackers and regular deli aisle shoppers. The project consisted of an optimisation phase and a measurement phase.

The optimisation survey was conducted from 29 April to 4 May, and was designed to identify the highest-performing concepts, while highlighting any variations in preference by key subgroups or segments. Based on this exercise, Affinnova could identify the key choice drivers among attributes such as emotional attitudes, category insights, product benefits, calories and imagery. In the unlikely case that a leading concept might end up being unfeasible for any reason, analysis also evaluated the strength of alternative variants and the relative risk of substitution.

In the measurement phase, respondents evaluated the top five concepts to emerge from the optimisation phase along with an existing Pretzel Crisps product and four leading competitors. The survey administered both a sequential monadic test and a discrete choice test. This hybrid approach enabled Snack Factory to assess the breadth and strength of appeal of these ten products and concepts. Key measures included purchase interest, ‘uniqueness’ and believability. Additionally, the survey provided for qualitative consumer feedback on the top concepts, including likes, dislikes and areas for improvement.

The results of both phases of research were delivered on 17 May so that Snack Factory had time to prepare marketing materials for the 5 June introduction of the product at the IDDBA conference.

The findings
The top concepts that emerged from the optimisation phase were primarily differentiated by flavour, which was the strongest driver of consumer choice. Surprisingly, given that fewer than half the flavour varieties included cheese, four out of the five top concepts included cheese, making it an important element of a bold and spicy offering.

Packaging had some impact on choice and had important interaction effects with the flavours. Existing purchasers of Pretzel Crisps had a stronger affinity for concepts that emulated the current packaging than non-purchasers did.

The measurement phase showed that three of the five top concepts beat the leading competitor and two showed the potential to outperform Snack Factory’s existing Buffalo Wing variety. The top concepts revealed different levels of preference by subgroup, with the jalapeño flavour offering the greatest potential to win over people who were not currently Pretzel Crisps buyers.

The outcome
Based on the study results, Snack Factory moved quickly to introduce two new flavours to its bold and spicy category: Chipotle Cheddar (“unique smoky flavour of chipotle peppers, robust creamy cheddar cheese and zesty spices”) and Jalapeño Jack (“combining naturally spicy jalapeño flavours with creamy, mildly tangy Monterey Jack cheese”). It was able to introduce both new flavours in time for the 5 June conference. The company also repackaged its existing Buffalo Wing variety to be consistent with the packaging style introduced for the two new flavours.

Grocery stores and independent delis were favourably impressed with the new flavours, leading to wide distribution from the outset. “This unique approach to consumer research helps to put metrics behind the emotional task of determining which new products to offer,” said Snack Factory’s Perry Abbenante. “We explored all of our options and I had greater confidence to approach the trade with our new products, as we have greater visibility into how they will perform at retail.”

A starting point of 182 concepts is small by Affinnova’s standards, so studies of future Snack Factory product offerings will be even larger.

The early sales outlook for the new flavours is promising, set to outpace the initial sales rate of Buffalo Wing Pretzel Crisps, which were a big success the year before. Preliminary results from the field confirm that the new flavours are complementary to the Buffalo Wing offering and also capture market share from leading competitors, making Chipotle Cheddar and Jalapeño Jack great additions to the Pretzel Crisps product portfolio.