Defending against private-label

Editor’s note: Randy Brooks is president, Bob Kushner is research consultant, and Aileen Beatty is research consultant, at Directions Research, Inc., Cincinnati.

A brief review of the trade press over the past five years or so turns up article after article and reference after reference spotlighting the growth of private-label offerings at the expense of national brands. A brief summary of the number, tone and content can be seen from the following:

  • A Google search of the Internet for all references to “private label” turns up 1,140,000 hits.
  • The number of categories with a private-label presence is growing. According to ACNielsen, since 1997 the percentage of categories with a private-label presence has grown from 69 percent to currently 75 percent.1
  • Private-label items are taking share from manufacturer brands. During the 1990s, sales of store-brand products in supermarkets increased from approximately 13.5 percent of store sales to nearly 16 percent.2
  • Private-label share in numerous consumer product categories is substantial and growing. The price differential during the time spanning 2002/2003 is very significant as the table shows.3

The growth of private-label products is frequently viewed as a threat to brands, leading some pundits to ask: are brands dead? At the same time retailers are viewed as increasingly powerful, with slotting allowances and a management focus on trade relations at the top of the agenda at most CPG manufacturers more often than brand management and advertising. In addition, national brand manufacturers (many of whom manufacture private-label products) are under significant pressure from retailers to reduce the costs of their branded offerings to support “lower everyday price” positions or provide extensive levels of trade dealing/price promotions.

A study sponsored by the Private Label Marketing Association conducted in September 2000 claimed that approximately seven of 10 U.S. supermarket consumers consider the quality of private-label products to be the same or better than national brands.4

How did we get to this juncture? How is it that many times branded products do not appear to offer adequately higher levels of quality to support the price differential that exists? But more importantly, how can CPG firms effectively defend themselves from this strong competitor?

Cost reduction research

Product testing has been an intensive activity at CPG firms for the past 50 years. Tests have always been done for a wide variety of reasons including: new product concept/usage, flavor line extension, product improvements, and cost reductions.

In the past 10 years or so the frequency of cost reduction tests has grown significantly. The appeal of this strategy is obvious. A cost reduction program can immediately put money on the bottom line. In addition, a lower cost of production can support more aggressive pricing, trade dealing and support for the brand. The cost reduction strategy is of course: “Reduce the cost of a key brand to save millions of dollars, but not jeopardize the brand’s health. A reduced-cost product that is invisible to consumers could be a home run for the firm.”

The idea of intentionally degrading powerful brands is clearly not risk-free nor lacking in controversy. Indicative of the degree of discomfort such initiatives have created, many firms now label cost reduction projects with politically correct names such as: cost containment, margin enhancement, efficiency, squeeze.

Our firm has been called upon to conduct tests of many product offerings - both our clients’ offerings and a wide array of competitive products across most major consumer products categories. A database of these tests called Navigator has been created to enable a cross-category/method analysis of results.

Design decisions made in cost reduction cases can entail huge risks to the health of key brands in a firm’s portfolio. We have analyzed the Navigator database to see if we could use data to assist in the design of cost reduction tests that are safer - that is to say, more sensitive to subtle differences that might exist between the current formulation and the cost-reduced version. In a number of areas we have uncovered findings that, if used, would add a degree of confidence that the cost-reduced version is, if it passes the test, unlikely to erode consumer satisfaction with the new, less costly version.

Cost reduction test designs

When beginning a major cost-reduction initiative, decisions must be made that can have a profound effect on the sensitivity of the design, including:

  • Blind or branded - Should the product shown to respondents have any identifying labels or logos?
  • Test environment - What is the physical setting in which respondents will evaluate the product?
  • User qualifications - What type of respondent do you want evaluating your new product?
  • Sensitivity of design - How sensitive do you want the design to be to detecting changes in respondent opinion?
  • Decision rule - What amount of difference in ratings between the original and new product do you consider acceptable?

Blind or branded

Consider the large amount of information that is given to a consumer when a branded test is conducted:

  • The brand’s image in terms of past marketing and advertising, any positioning and perception of who uses the brand.
  • The perception of price and, more importantly, the perceived value.
  • Experience with the brand itself (both good and bad) by the respondent.
  • The product performance in terms of its overall quality, durability and reliability.

The Navigator database includes nearly a dozen pairs of tests where alternative formulations were tested both blind and branded. In virtually every case, the degree of difference noticed by consumers was substantially less when the brand name was present.

The argument made for branded testing is that the equity of the brand is an inseparable part of the product,  so shouldn’t it be considered? While true, the inescapable conclusion is that subtle damage can be done to the product if the test is only done branded. A test conducted in such a way is less sensitive and less likely to identify a difference when one exists. One option is to conduct both a blind and a branded cell when conducting the test. The blind cell will help to examine the pure product effect while the branded cell will help to determine the ability of the brand name to withstand the change.

Test environment

The primary venues for product testing are in-home and CLT (central location test). In-home tests entail providing the respondents a sample of the product to use at their home while CLTs involve recruiting respondents to visit a central testing facility to evaluate the products.

Both have advantages given certain research objectives. Central location tests are, generally, the more preferred method for cost-reduction as the respondents are more able to focus on subtle product-to-product differences due to the temporary removal of everyday life distractions.

Central location tests can simulate the home-use experience through asking the respondent to use the product in situations and realistic tasks that mimic those experienced at home. Important information can also be observed during a CLT such as how a respondent opens, prepares and dishes a normal serving, all points in the product-use process that can be affected by cost-reduction.

Cost and timing effects also favor consideration of CLTs for cost-reduction research. Many times, CLTs provide a lower cost and quicker turnaround while providing better control of the test by ensuring a consistent test environment.

Despite the advantages presented for cost-reduction, CLTs do not work well for all categories. Some products (many personal care products for example) simply must be tested in-home to obtain realistic reactions from the respondents.

User qualifications

Different user groups will likely have different perceptions of the new product. Light category users will be less sensitive to changes than heavy category users. Heavy users of your brand will likely have different perceptions of the new product than light users. Given this, the analysis of cost reduction research data should involve an investigation of user group differences.

Sensitivity of design

By definition, the change to be tested in most cost reduction research is not very noticeable (the objective usually is that the difference is not noticeable at all!). However, the test design should be adequately sensitive to detect if there is any shift in product satisfaction.

Many designs are used in cost reduction research projects, the most basic being the monadic design (one cell of respondents sees the current product and another cell sees the test product). This is not very sensitive, as a respondent is exposed to only one product variant. The use of a sequential monadic design allows the detection of product changes as the respondent sees both the current and test product. This has the benefit of allowing a built-in point of reference to evaluate the products.

A sequential monadic design also allows for more sensitive statistical testing. A matched group provides greater sensitivity by accounting for the respondent-to-respondent portion of the overall variance. Some commonly used tests that do this are the matched-groups t-test or repeated measures analysis of variance (for more than two products).

Another analytical benefit of a sequential monadic test involves the analysis of the first/second position results. An evaluation of 109 paired sequential tests in the Navigator database reveals the number of significant differences between products is greater in second position than in first (see Figure 1). In other words, an evaluation performed within the context of a previous product allows for greater sensitivity.

Decision rule

Occasionally, marketers attempt to duplicate a successful cost reduction program to further decrease product costs. While on the surface this sounds like an effective plan, this course has a serious flaw: the repeated use of a decision rule that allows a potentially inferior product to eventually replace the original.

A decision rule typically used in cost reduction research is, “Adopt the margin-enhanced product only if it is statistically at parity (could be below but not much), or it is equal to/directionally ahead of the current.”

Figure 2 demonstrates that if an 8 percent difference in product preference is acceptable for the first test, this translates into a maximum of 92 percent likely preferring the Test 2 variant. Applying the 54/46 decision rule means the Test 2 variant is actually different from the original by a 42 percent to 58 percent margin. If this continues, after four waves, the Test 4 variant is actually inferior to the original product by a 36 percent to 64 percent margin.

This subtle decay in the product (referred to as a “stair-step to disaster”) could become noticeable over time and be perceived as a slow degradation of product quality. To avoid this, comparable ratings to the original can be accurately maintained if a decision rule of 50 percent/50 percent is kept throughout the process, thus not allowing seemingly small differences to accumulate.

Extreme care

Private-label products have a growing impact and are likely to continue to exert pressure on the sales and vitality of branded products in many categories. When manufacturers look to cost reductions as a means of competing with private-label offerings, testing must be done with extreme care and a focus on sensitivity, i.e., a central-location, blind-product, and sequential monadic design. The research should be conducted to insure that the quality of the brand is not undermined due to the use of methods that could mask the impact on consumer satisfaction of the cost-optimized formulation. 

References

1 U.S. Trends in Private-label, ACNielsen U.S., September 2003.

2 German, Gene. “Are Consumers Buying More Private-label (or Store Brand) Products?” Department of Applied Economics and Management, Cornell University, June 2002.

3 The Power of Private-label: A Review of Growth Trends Around the World, ACNielsen Global Services, July 2003.

4 Cuellar, Sandra. “Private-label Brands – A Growth Opportunity for Retailers and Produce Suppliers.” Food Industry Management Program, Cornell University, February 2002.