It goes without saying that customers are the lifeblood of any business. So understanding your customers’ preferences—and, indeed, the preferences of the people who are not your customers—is essential to the process of developing, marketing and, ultimately, selling any cosmetic or toiletry brand. Given the generally accepted 80% failure rate for all new products, it seems apparent that some companies are leaving much to chance when it comes to understanding their customers, and as is frequently the case in many games of chance, the odds are stacked against the player.
However, appropriate and properly conducted consumer studies can tip the odds in favor of the manufacturer. While they can not guarantee the success of a new product, they can identify key attributes of a product that might motivate a potential customer to buy and highlight the attributes that might prevent a one-time customer from buying again. These studies can demonstrate whether consumer perceptions match the development objectives and potentially expose whether the objectives were desirable in the first place. Armed with this information, brand managers can work during the new product development process to include more of the desirable characteristics and also direct the marketing to focus on the aspects of the product that are most likely to motivate customers to purchase.
There are many types of consumer studies that might be undertaken. However, it is important to consider all aspects of the product experience when designing the work. One would not expect to be properly able to evaluate a bath soak or shower gel simply by sniffing the bottle in the middle of a busy shopping center, even though that is where the first purchase is likely to be made. Rather, the ideal scenario for assessing consumer perceptions of products is to match the situations of use as closely as possible, and that means conducting in-home trials. That said, there are specific consumer studies for which trials based only in-home are not the appropriate forum. Sensory evaluation, cognitive studies and dermatological studies, for example, will also require an element of a laboratory-based approach. However, these studies are often directed at answering different types of questions, which might ultimately be allied to the results of in-home trials to form a more complete picture. Neither should this preclude linking the results from in-home trials with other instrumental or mechanical measures of the product. This type of statistical analysis can often create a bridge between what the consumer experiences and how the material being assessed is formulated.
The essential point about any consumer study is to define the questions the study is intended to answer in advance. It is an obvious observation to make, but still worth making since a study designed to answer a question such as, “How does my product compare with brands A and B?” will require a different design from the study that asks, “What should I change about my product to win market share from brands A and B?” Even though the ultimate goal might appear to be the same in both cases, the specific questions asked of consumers might be different, and indeed the profile of consumers selected for the study might also be different.
Defining the study from a commercial perspective is one thing. Defining the study from a legal perspective can sometimes be more difficult, especially when considering cosmetics and toiletry products that sit in the grey area between cosmetics and pharmaceuticals. Here, one has to question at what point does a consumer study have to be considered as a clinical trial, or at least to question whether best practice dictates that some aspects of the study should be carried out with the same rigor and controls as a clinical trial.
This is not merely a matter of semantics.
The World Medical Association Declaration
of Helsinki (Tokyo 2004) states that, “Medical research involving human subjects should only be conducted if the importance of the objective outweighs the inherent risks and burdens to the subject. This is especially important when the human subjects are healthy volunteers.” In Europe, European Directive 76/768/EEC and its amendments on cosmetics as well as European Directive 2001/20/EC on Good Clinical Practice (GCP) in Clinical Trials set the framework for addressing these issues. The Cosmetic Directive 76/768/EC defines a cosmetic product as any substance or preparation intended for placing in contact with the various external parts of the human body or with the teeth and mucous membranes of the oral cavity with the exclusive or principal intention of cleaning them, perfuming or protecting them in order to keep them in good condition, change their appearance or correct body odors. Any “cosmetic” product that seeks to claim to do more than clean, perfume, maintain condition, change appearance or correct body odors might need to be considered as something other than a cosmetic.
In any event, before any product can be used in a consumer preference study, it must satisfy safety criteria. This usually means it must be supported by toxicological data on its ingredients, results of animal testing and any results of prior use on humans.
Selecting consumers to participate in in-home trials is a relatively straightforward matter, and a group of between 150–200 volunteers is generally considered statistically viable. RSSL’s database of volunteers includes approximately 5,000 households, and there are many agencies around the world with equal capacity to call on volunteers of every conceivable gender, age, ethnic, economic, social and religious mix. Some agencies will hold basic data so that it is possible to query a database to narrow down the target range to, for example, women in the 30–45 age group. However, depending on the product to be tested, it will be necessary to carry out a further interrogation of this target group, perhaps to identify those consumers already using similar products habitually or users of specific brands.
It is possible, and sometimes necessary, to be highly prescriptive about the profile of the consumers chosen for the study so that if the target market is highly defined, the consumer group can be equally and appropriately highly defined. One must exercise a little caution in being too defined, of course, since it might be a mistake to define a target market too narrowly. There might be many reasons why an older male will choose to purchase a product that is targeted at a younger, more athletic market. Similarly, a female might purchase a product for other members of her family, which the partner or child might never think to purchase for themselves. On the other hand, it is obvious that recruiting a large sample of bald men would be inappropriate for a consumer study of a shampoo or hair colorant.
In short, the ultimate definition of appropriate criteria for inclusion in an in-home consumer panel will depend on what questions are to be asked and on the profile of the target market.
In addition to the commercial issues that the study is designed to address, there are ethical and, occasionally, medical considerations that might need to be taken into account. The safety of the product itself was dealt with above, but even where no medical implications are involved, each participant needs to give informed consent to take part in the study. To ensure informed consent is given, it is both courteous and good practice to issue each participant a detailed explanation of the study and what will be expected of all parties. The National Research Ethics Service has published a set of guidelines covering the points that need to be addressed in providing information to participants and the completion of consent forms.
It is important to make sure that a participant’s consent to take part in the trial includes their consent for the data arising from their participation to be used and made available to the sponsor, auditor and regulatory authorities, where appropriate. This is especially necessary in countries where there is data protection legislation, but in any event, it is advisable to protect against some later claim by ensuring that the consent form is signed.
Clearly, the research agency managing the study also has responsibilities to protect personal information given by participants, as well as to protect the health of participants during the course of any study—as a matter of course in the U.K., for example, under the requirements of the Data Protection Act. Where the product might intend to claim some therapeutic benefit or where participants might have some relevant medical condition—psoriasis, for example, in a skin care product test—it would be advisable to include a medical practitioner on the study team and also to require a participants’ general practitioners be informed of their participation in the study.
Selecting Samples for Testing
Careful consideration must be given to the products that are to be assessed and there are two elements to this choice. The first is a consideration of the material that is being tested, taking in all of the factors involved in product usage. This is likely to include the number of use occasions that one should expect a consumer to require in order to result in a realistic assessment of the product. A shampoo may require three or four uses, for instance, while it may be possible to evaluate the acceptance of the fragrance of a body spray over a relatively short period of time. The longer a consumer is involved in a product assessment, the greater the risk of fatigue and boredom. The consequence of such a situation is that while the consumer will continue to provide an answer to the question posed, those answers become less reliable as the length of trial increases.
Once an appropriate number of products for trial has been established, the next critical element is deciding which products will, in fact, be tested. In a multi-disciplinary project team, there will inevitably be a number of different opinions as to which products should be tested, all equally valid. The discussions and ultimate product choice should ideally be based upon a marriage of the needs of the commercial function (competitor products) and the product development function (how all the products are different from one another). An important guiding principle here is that consumers will find it difficult, in the context of a blind assessment, to discriminate between products that are very similar to one another. Therefore, if there are multiple competitive products that are very similar functionally, it is advisable to eliminate all but one, assuming the results of the one will serve as proxy for those eliminated, allowing more focus on products that offer different properties. This will have the effect of broadening the range of product attributes the consumer is exposed to, which, in turn, enhances the overall value of the piece of work.
Control of the consumer experience of the product is clearly a critical part of the design of any experiment. In addition to the provision of clear instruction for usage, it also incorporates elements such as the order in which products are assessed and how many products a consumer is asked to assess as part of the trial.
The order of presentation should be varied—ideally following a statistically controlled design in order to minimize any bias engendered by the order of trial. Human nature is such that it is most likely that the first product in a sequence will be assessed on its own merits, whereas the second is more likely to be assessed in comparison to the first one tried, and so on. Therefore, it is important to randomize the order in which all products are seen, thereby spreading this effect throughout the population.
Preparing the Samples
The guiding principle behind the preparation of samples is always to prepare more than you think you are going to need. Within the context of a consumer test, the cost of product is relatively small in comparison to not being able to do everything that you would like simply because you do not have enough product.
Unless it is important for participants to know which product they are evaluating or unless they are also being asked to evaluate aspects of packaging design/performance, the samples should be provided blind. This is the simplest way to ensure that participants are only influenced by the product, and not by any assumed properties derived from a brand name or previous marketing messages.
Samples are usually assigned a blinding code, and should be distributed in a controlled fashion with clear instructions about how the product should be used. This might be an instruction of when to use it, how much to use at a time, whether to smell the product in the packaging and/or on the skin, whether to assess the color in daylight or under artificial light, and so on.
Clearly, there is an element of trust involved, and there is no sensible means of assessing whether the instructions have been followed. Retrieving the samples once the study period has elapsed allows measurements to be made of the amount of product used, but even here, there’s no absolute guarantee that the participant has actually used the product.
That said, it is seldom the case that the rewards for participating are such that anyone would volunteer purely for financial gain without intending to assess the product in the manner required.
Designing the Questionnaire
The ideal questionnaire to send to consumers, along with the product samples, is likely to be comprised of a combination of scaled responses to closed questions designed to evaluate the participants degree of like/dislike of key product attributes. These might be supported by open questions that invite consumers to elaborate on their reasons for liking/disliking aspects of the product. It is preferable to keep the questions simple and to limit the number to no more than 20. Any more can overwhelm the participant and is likely to result in less reliable responses.
A paper-based questionnaire is still the format most often used for gathering responses, but electronic media offers an option that is becoming increasingly widely used. Internet forms, e-mail and text messaging are options that might be considered. For this aspect of testing, too, there is an element of trust involved—notably that participants record their observations within the prescribed time scale after using the product and, of course, that they record their conclusions accurately.
Analysis of Data/The Final Word
Data analysis is a large topic that justifies an article in its own right. Suffice to say, there are many different types of analysis that could be conducted, and care should be taken to ensure that the most appropriate one is selected, in the light of the context of both the overall test protocol and the specific objectives of the work.
In–home studies of cosmetic and personal care products provide wide ranging data that can be used to better understand consumer preferences—guiding changes in formulation of new or existing products, possibly, and to direct the marketing strategy of the product in the marketplace. However, the quality of the data depends on the quality of the study.
Valerie Hart SCS Dip., is the manager of RSSL’s clinical sciences department, managing a wide range of safety and efficacy research studies for clients in the pharmaceutical, OTC and cosmetics and toiletries industries. She has considerable experience of conducting phase I and phase II clinical trials.