In 2020, data dominates.
That was one of the key themes from WWD’s Digital Beauty Forum, held in February in New York City, where a handful of companies focused on how they use data in order to make and market products to consumers and provide services to beauty companies.
These days, beauty brands are using all kinds of consumer data for things like product development — Olay’s new Retinol 24 product is one example — and marketing, think data-supported decision-making for social media images.
Olay is one key proponent of using data to inform product development, but took a winding road to get there, said Eric Gruen, North America brand director for the skin- and body-care brand, which is owned by Procter & Gamble.
The brand did well between 2000 and 2010, but that year, started experiencing declines as it failed to maintain relevance with the direction the market was moving in. “[Consumers] would call us, “Oil of Old Lady,'” Gruen said. “When we would try to get people to work on our brand within the P&G family, they would look at you like you have three heads.”
It took a while — Olay’s decline continued through 2016 — but gradually, Olay started implementing digital strategies into its workflow and developed a deeper understanding of what consumers are looking for from skin-care routines. Eventually, the brand launched Olay Skin Advisor, a web-based tool meant to help people understand the right products for their skin.
The tool also enables P&G to collect data, which helps inform the brand’s strategy. Users take a selfie, which goes into a “smart algorithm,” and then people are asked between five and seven questions about their skin, before the app reveals the consumer’s skin age. Gruen referred to it as the “game-ification moment” and noted that’s where the brand sees a lot of engagement.
The final step is product recommendations customized for each user.
Several years after launch, the Olay Skin Advisor has collected more than six million data points, he said, which have significantly impacted product development.
For example, Olay learned that a big percentage of consumers actually desire fragrance-free skin-care products — something that wasn’t even on Olay’s product development roadmap.
“We had a quick moment where we were like, ‘uh oh.’ We’ve got to come out with fragrance free, because we’re not satisfying the consumer and what she’s telling us she wants,” Gruen said.
So Olay launched scent-free versions of its Olay Whip facial moisturizers, which have subsequently become “just as big” as the original, Gruen noted.
“This is the first time we took those data points, those consumer engagements, and actually transitioned our mind space from a digital-to-sell mind-set to a digital-to-design mind-set,” Gruen said.
Olay has continued the strategy for future products, including Olay Retinol 24. The brand saw the retinol was the number-one searched ingredient, and decided to create a new product.
“We took this data and said, how can we bring to life, and how fast can we bring to life, a new retinol product?” Gruen said. Then, the brand highlighted the word “retinol” on the products packaging, making it about the same size as the Olay brand name, a move that took a lot of convincing internally, Gruen said.
“Trying to convince people that your ingredient should be the same size as your brand market is not easy,” he noted. “Consumers were saying, ‘I need a brand I can trust with an ingredient I want…the reality is, they’re buying it because they know it has retinol.” With the fragrance insights from the prior practice, Olay only launched Retinol 24 in a fragrance-free version.
Over at Prose, a custom hair-care business, data is an integral part of the operation.
Prose customers take a 25-question online quiz in order to determine their best custom formulations and product regimens. Questions range from hair type and scalp conditions to environment and lifestyle habits — Paul Michaux, Prose cofounder, said that quiz leads to 135 data points that the company can use.
“We take this information, we put it in our algorithm that we created in-house, and that will create a bespoke formula just for you. It’ll also tell you which products … should be part of your routine,” Michaux said.
Products get shipped out to customers, who can provide additional information via Prose’s online Review and Refine feature. It also allows consumers to provide information on themselves — if they have a dietary change or move from city to another, for example, their shampoo formula may need to shift, as well.
“It gives us a massive amount of data every day from customers,” said Arnaud Plas, cofounder and chief executive officer. “By leveraging that, we’re continuously improving the product. It’s basically bringing the network effect that is well known in the social media space into product. It helps us to really continuously improve the satisfaction.”
“Your feedback is not only helping your formula, but also the formula of everyone else,” Michaux added.
Prose has a team of in-house data engineers, based in Paris, that aim to understand similarities between customers, and what ingredients can be tweaked in order to maximize satisfaction. The business says it has been able to increase satisfaction by 30 points, to about 90 percent, in the past year.
The data portion is also absolutely necessary for running a personalized product operation. Prose’s formulas are each custom made.
“What we’ve made is unique, but it’s a composition of ingredients and parts from a manufacturing perspective that were on the market,” Plas said. “Where we’ve been innovative is by combining AI to automation to building a full stack and software to power this and make it scalable.”
Ipsy, too, has used consumer data to create products, as well as to personalize its beauty subscription boxes. The business launched its first in-house brand last year, Complex Culture, based on information it has on consumers from its vast subscription business, cofounder and chief executive Jennifer Goldfarb said.
“We know we put something really quality out into the market. We’re going to continue to do that, but really selectively — we are using all of our data…we’re really targeting the whitespace where there is an unmet consumer need,” she said.
“From day one, we’ve invested in industry leading technology and data science,” Goldfarb said. “[Sending personalized beauty products] is a super hard problem, and the only way to do it well and do it at scale is to use technology. From the minute you join Ipsy and you take our beauty quiz, to the 160 million product reviews that are on our site, our team is collective signals that help us better understand you and what are the best products to send you every month.”
Eight years into business, Ipsy, which was co-founded by original beauty YouTuber Michelle Phan, has 25 million people engaged with the business, Goldfarb said, either as subscribers, at in-real-life activations, or online. The business has more than 300 employees and reached $500 million in revenue in 2019.
The key to its operations, and its data gathering, is building out a community.
“The next big idea in beauty is community,” Goldfarb said. “The brands that will win in the future are the ones that can activate their customers, their followers, their fans, and really bring them into their brand, help them build the brand, product development, marketing, all aspects of the business.”
Ipsy tends to gravitate towards YouTube for research, and there, found that consumers weren’t necessarily looking for “the best” products, but instead looking for the products most suited for themselves, Goldfarb said. That learning led to Ipsy’s Glam Bags. “Consumers were craving a brand that would listen to them and understand them,” Goldfarb said.
“How do you know what someone wants? Why don’t you ask them,” she said.
Through technology, Perfect Corp., which develops AI and AR programs for beauty companies, is aiming to provide product recommendations to people without necessarily having to ask them what they want. Wayne Liu, general manager of Perfect Corp., said it uses data from its 250 beauty brand partners to help make product recommendations to consumers, for both makeup and skin care.
Liu contends Perfect Corp.’s offerings help drive sales. “Alibaba actually whispered to us, just six months into their AR, it increased 4x conversion rate,” he said.
“The technology is trying to solve the most difficult problem in the world [and] offer personalized recommendations. The reason I say this is difficult is because we’re trying to figure out what girls want — that’s the very difficult part,” Liu joked.
Perfect Corp. has data from companies and has users taking selfies in apps. It can then identify a person’s facial attributes, recommend makeup, and show a user what that suggested makeup would look like on them. The service also applies to skin care. “The data needs to be brand agnostic,” Liu said.
In terms of using data to inform social media practices, Michelle Belcic, the vice president of brand strategy at Dash Hudson, said the company has the technology to help. She asked the audience a series of questions while comparing potential social media posts — which one would have higher engagement?
For example, there was an influencer-in-the-wild post for Revolve, versus a posed outfit show on a white backdrop and an earring shot versus a necklace moment — the influencer-in-the-wild and earring photos were the winners.
For the most part, the beauty audience did okay — but Belcic contends there is a science to being able to determine which images will thrive on social, and which will not. “The number one most important factor in driving sales lift from any digital campaign is the quality of the creative,” Belcic said.
“We can tap into all the different elements of what’s in the photo…the tone, the field, the patterns of the pixel and tie these thousands of visual cues back to historical engagement,” Belcic said. “You don’t have to guess anymore.”
She added that beauty brands have common misconceptions when they are approaching social media that what works for other brands will work for them, and that what performs today will also perform tomorrow. For Goody, the hair accessories brand, quirky quotes and interesting braids perform well, Belcic noted.
“Visual preferences change, so we have to have that real-time data so as our audience changes and evolves, we have that and can tap into that as well,” Belcic said.
For more from WWD.com, see: