When you buy beauty products, the last thing you want to experience is an allergic reaction of some sort. Fortunately, many stores offer testers, allowing you to apply small doses and see what happens before committing.
However, for some products, that’s not a realistic option. The FDA’s investigation into hair care products by the celebrity stylist-founded manufacturer WEN found that they caused scalp irritation and hair loss. Moreover, in the wake of the pandemic, companies are facing intense scrutiny over the unhygienic practices surrounding in-store testing.
Concerns over cosmetic safety have led to explosive growth in the market for ‘clean’ beauty products. When you buy all-natural anti-dandruff shampoo, you’re not only seeking efficacy but also gentle, hypo-allergenic care.
But the real solution to a safer industry for all lies in the way we use big data.
The root of the problem
The movement towards clean cosmetics was spurred by growing concerns about the presence of potentially harmful products, like BPA and phthalates, in stuff we apply to our skin and hair.
Celebrity-endorsed brands like Gywneth Paltrow’s Goop and Jessica Alba’s Honest Company have crested the ensuing wave of consumer sentiment. And big brands like Sephora and Unilever have taken steps to present a cleaner, more transparent image.
But the truth is that every industry player has the potential to slip up, commit negligence, or otherwise engage in substandard testing practices.
At the heart of the problem is the lack of regulatory control within the industry. In essence, the FDA requires beauty companies to conduct testing but leaves the specifics up to their discretion.
Clean beauty ingredients may be safer in theory, but if the manufacturer doesn’t conduct rigorous testing, the final formulation may trigger irritation or allergic reactions. And professional dermatologists may recommend nonallergenic products such as petrolatum that end up wrongly demonized by retailers striving for a clean image.
Thus, cosmetic producers’ attempts to go green may come from good intentions but could end up causing harm to individual consumers anyway. And that’s bad news for all industry stakeholders in the long term.
Facing the limitations of testing
The FDA doesn’t intervene in the cosmetic industry because except in specific cases, such as when color additives are involved, it has no legal mandate to do so. For change to be initiated in this quarter, federal law must be rewritten by the US Congress.
That possibility isn’t out of the question, but it’s also not likely. A more promising approach would be to keep the industry self-regulating but improve testing protocols.
There are significant barriers to that happening. People are highly diverse in terms of genetics and physiology. Lifestyle factors can also affect our reactions to cosmetic ingredients. The result is that testing a small sample size might not catch every averse outcome.
On the other hand, testing more people is expensive and time-consuming. Consider that companies also have to test for long-term product stability, anti-microbial preservation, and compatibility with packaging. A lengthy and rigorous testing process can increase costs to the point where smaller companies may find them prohibitive.
Animal testing may have been used before as a cheap alternative but has since been banned in many countries out of ethical concerns. And even if manufacturers try to go this route in countries where permitted, reactions vary even more greatly between species than between individual humans.
The data solution
The solution might be one that many companies are already using extensively but geared towards an entirely different purpose.
Big data has been widely used for years now in the beauty industry to obtain insights on consumer behavior and improve marketing reach. But analytics can also be used to crunch the numbers with respect to product safety and health concerns.
Some companies have already taken steps in this direction. Proven Skincare, a beauty startup, uses an algorithm formulated by Stanford scientists to distill personalized product recommendations from the vast amount of information available across the internet. The company says that its machine learning capabilities include processing product reviews, peer-reviewed scientific articles, and a database of over 20,000 cosmetic ingredients.
Another startup, Alcheme from Singapore, explores a different way to harness the power of AI. It allows users to submit selfies which their algorithm processes in reference to its internal database of images, along with answers to a questionnaire. The AI then submits an analysis of their skin conditions, along with a list of recommended ingredients that will then be used to create a bespoke product for them.
These innovations point to a potential path forward, and any beauty company should consider similar data-driven routes to ensure better product safety and consumer trust.
