The “Science” of Cosmetic Product Testing

There was an interesting article about ‘clinical testing’ of cosmetic products. Essentially, the authors suggest that the science used to support the claims made by the nutrition supplement product Inneov Sun Sensitivity is faulty, biased, and practically worthless.

I don’t yet have access to the original BJD article but if the news report is accurate, indeed this is sketchy science. The extraordinary claims are not quite supported by the testing.

Unfortunately, this is usually the case with studies designed to prove a point rather than discover a truth. In the cosmetic industry, it is called Claims Testing.

Let’s look at some of the questionable testing that goes on in the cosmetic industry and suggest how it could be improved.

Using consumer opinions to make claims

This happens all the time. When a cosmetic manufacturer wants to make a strong claim their product, they go to consumers and see what they think. If it is a bad study, consumers will be given a product that they use for a certain amount of time, then they answer questions about it. Their answers will then be used to support claims made about the product.

Why is this questionable?
There are a variety of reasons this is not good science but the biggest is that consumers are easily fooled. If they like how something smells or feels, or they just enjoyed the overall experience, they will be inclined to rate everything higher. For example, if a consumer likes the fragrance of a hair conditioner, they’ll be much more likely to rate highly the performance of the product for shine, detangling, manageability, etc. The positive feelings about the fragrance has a “halo” effect on all other attributes.

So, if you ask a consumer whether a product made their hair more shiny, the answer will likely be influenced by something that has nothing to do with shine at all!

How to make it better.
If you are really interested to know whether a product improves a specific attribute consumer testing isn’t the best thing to use. Better are lab tests or trained panel tests. However, if you want to use consumer tests then at the very least make it a double blind controlled test. This means that you give them your test product and a control product that look and smell the same. You should then have the products made by someone who is unaware of the test purpose so you don’t know which samples are which. You should also use a minimum of 30 people for the test however, >100 is even better.

Using improper controls

This is a fairly common occurrence in the cosmetic industry. Perhaps most common is in studies done by cosmetic raw material suppliers. They show impressive looking data that is much less so when you consider the control. Frequently, a raw material will be evaluated against a ‘no-treatment’ control. It should come as no surprise that any skin lotion will improve skin condition when compared to no treatment.

Why is it questionable
The problem with improper controls is that you set up a test in which you know your product isn’t going to fail. You already know there will be a positive outcome. This is not science.

How can you make it better?
The way to make this testing better is to use a proper control. Not only should you have a positive control, but you should have a negative control. Suppose you want to know if your moisturizing shampoo makes hair easier to comb. Your proper controls would be a Normal shampoo and an alternative Moisturizing shampoo. Compare the performance of all 3 to answer whether your product is indeed better than other moisturizing products.

Using unrealistic lab tests to support claims

To make incredible claims, products are often taken into the lab and tested in ways that specifically support the claim. For example, to measure hair shine a single fiber is sometimes used. This is fine if the claim is that it makes a hair fiber more shiny but it doesn’t answer the question whether it makes bundles of hair more shiny. Similarly, the thickness of a hair fiber is measured using a laser micrometer. This is supposed to support “thicker hair” claims even though the increase in thickness is something consumers would never be able to notice.

Why is it questionable
The reason that these tests are questionable is that they create support for claims that are different than the consumers might believe. If a product claims to make hair 50% more shiny and the data is based on single fiber tests, it really isn’t true.

How can you make it better?
Lab tests like these are excellent to use while developing new products and refining prototypes. But you can’t just stop with this testing. Additional testing should be used to check to see whether the changes are noticeable to a trained panel and ultimately a consumer.

There is some excellent testing going on in the cosmetic industry but it could be improved. Using proper controls, minimizing consumer halo effects, and validating lab tests with trained panels are simple measures that could lead to significant improvements.