Renude hosted a webinar alongside Catherine Bossom, Founder of Yellow Flamingo and CEW Board Member, discussing the topic of AI in skincare. During this conversation, we discussed;
- Why the skincare shopping experience is so overwhelming
- What the impact of this overwhelm is for brands and retailers
- How Renude combines Computer Vision, Machine Learning and Large Language models to offer customers personalised skincare advice
- Our new launch: AI Skin Chat
- The commercial uplift that these tools can provide to brands and retailers
Read our summary below, or watch the webinar in full here:
Why is skincare shopping so overwhelming?
The skin is a complex organ, many factors influence its functioning and its constantly regenerating. This makes it complicated to understand exactly what it needs, and whether the products you’re using are working (as results take time.)
On top of this, there is a large range of brands on the market, with new products being launched every day. There has also been an increase in interest in ingredient science, resulting in a lot of content around this, with new ingredients being highlighted as the next best thing.
This combination of challenges to understand our own skin, and an excess in options makes it very difficult to choose the right products. One of the most common ways people buy new products is a recommendation from a friend, or an influencer they trust. But said friend or influencer is unlikely to have the same skin, and therefore their favourite products might not end up being yours.
What the impact of this overwhelm is for brands and retailers
The impact of this confusion can be seen in high abandoned basket rates, which average at 70-75% for beauty, some sources claiming up to 87% on mobile devices.
Beyond this, a poor match will lead to a negative product experience, and subsequently a bad review. Reviews are a critical step in the customer journey, so poor reviews can have a significant impact on future sales.
This overwhelm has also led to a level of mistrust within the industry. Customers will only buy from companies they trust, so gaining and keeping that trust is crucial for retaining customers and building Lifetime Value.
How Renude combines Computer Vision, Machine Learning and Large Language models to offer customers personalised skincare advice
AI is really good at a variety of tasks. Computer Vision (a field of AI) has proven to outperform humans in visual recognition tasks, this has been researched within the Dermatology field. This makes visual skin analysis a perfect application for Computer Vision. We can train AI with more image examples than a human Dermatologist could see in a career, allowing us to train in higher accuracy than a single human expert.
Machine learning is a well established technology for recommendations. You will likely interact with examples of this daily via Netflix or Spotify. In the case of skincare, Machine Learning is built using inputs (a person's skin profile, made up of their images and questionnaire) and outputs (their recommended routine), then designing the links between the two. With enough examples, it gets very good at creating skincare recommendations for new customers.
Large language models have advanced tremendously in recent years, opening new opportunities to provide conversational advice. The tricky part is the information source they are pulling from - some have high levels of inaccuracy (the internet) and so caution should be taken. As Renude already has half a decade of 1-1 customer skin advice from licensed aestheticians, we have a highly accurate vetted knowledge source to use in combination with conversation AI capabilities.
Renude’s new launch: AI Skin Chat
We revealed our latest launch during this webinar, an AI Skin Chat which combines all three of the above AI technologies to provide personalised advice to any skincare-related query.
Examples could be:
- Which moisturiser is best for me?
- I’ve seen this Vitamin C serum online, is it right for me?
- Does this product contain fragrance?
- Can you find me a product similar to X?
The options are endless, reflecting the huge variety in customer journeys that we see in skincare. Some customers identify a product through a friend or social media, then set out to research if it’s right for them. Others want to learn about ingredients before they purchase. This has historically taken consumers up to 90 minutes per product purchase, requiring multiple sources and touchpoints, often carried out over several days or weeks.
This advice can now be provided in one place, and that can be your website! Keep your customers on-site, help them to get their queries answered and build trust to retain them.
Commercial uplift using AI
75% of customers are willing to pay more for beauty products to get a personalised online experience.
Our hypothesis here is that because consumers have wasted so much money in the past on products that didn’t end up being a good match for them, they are reluctant to invest in another guess. Getting a personalised recommendation gives that customer higher confidence that a product will work for them, and as such de-risks the purchase.
What we have seen using our AI tools
- 2.5X uplift in AOV
- 2X uplift in basket size
- 40% increase in Average Unit Price
- 2X increase in conversion from click ‘Get Started’ vs. other site traffic
- 10X increase in conversion for those who receive a routine
Building Renude intelligence into your personalisation strategy
Renude’s AI tools are fully licensable, with API or plug-in options. Those hosted on Shopify/BigCommerce can be live within just a couple of weeks with no developer required, Renude manages the integration on your behalf.
If you’re looking to upgrade your personalisation strategy, connect with us via email to find out more.
