Pippa, can you tell us about your journey as a formulation chemist and what inspired you to co-found Renude?
I was obsessed with skincare from the age of around 14. Which feels totally normal now but it wasn’t 20 years ago, my friends all thought I was mad, religiously doing my Cleanse and Polish routine at sleepovers.
This curiosity on what made formulations different from each other led me to a degree in Cosmetic Science, and on to working in R&D for big brands like E45, No.7 and Veet, then on to high-growth startups like Beauty Pie where I was the 4th employee.
After a decade working in skincare, and being consistently asked for advice from those around me, I realised there was a huge gap in the market for a solution that helped people better understand their skin, and navigate the available options.
I started creating personalised skincare plans for those in my network, and as they started getting great results, word spread and I was completely inundated with these requests. I started exploring how technology could be used to scale such a service, when I very serendipitously met my co-founder Cate. She is a data scientist, and had also realised a need for this service after a close friend had struggled with cystic acne and rosacea and wasn’t getting effective support through medical pathways.
We teamed up to combine the worlds of skincare formulation and AI, working alongside a team of dermatologists and aestheticians to build-in skin expertise from all of these different specialisms.
What gap did you identify in the skincare industry that led to the creation of Renude and its AI technology?
Beauty is unique in that most products are purchased through a super wasteful trial and error process – our research shows that 62% of self-selected purchases aren’t a good match for the skin, and 28% are discarded before they’re even finished. This is costing the average UK woman £617 and 9 working days of research simply on the products that don’t end up being a good fit.
This is hugely wasteful, as well as being frustrating for consumers who often end up cycling through different products and in a lot of cases even making their skin worse.
Ultimately this is due to the lack of access to professional skin support. 69% of people choose to self-manage their skin conditions through over-the-counter skincare, so Renude’s technology provides a solution for these people to better understand their skin, and make confident purchasing decisions.
The result for the brands and retailers that we license our technology to is increased average order values, both through people buying more products and more expensive products (because they have confidence they will work for them), as well as increasing conversion rates and retention.
How does Renude’s new Skin Routine AI and Skin Chat AI technology work, and what makes it a game-changer in personalised skincare?
Renude’s AI technologies are designed to replicate the advice and recommendations that you would receive from a licensed professional.
To develop AI that can replicate a professional, you need to train it with thousands of examples of recommendations and advice made by these professionals. This type of data doesn’t exist, so our approach was to collect this directly, working with a team of licensed professionals to recommend products to consumers, and assess how the different combinations of products and ingredients impacted their skin. At scale, this dataset has allowed us to develop AI that can ‘think like an expert’ as essentially it has been trained by experts.
We have two products:
Our Skin Routine AI can analyse the skin for 32 different parameters, and recommend a complete, ingredient-led skincare routine. These routines consider the combination of ingredients and products to achieve the best results for the consumer.
Our Skin Chat AI is an industry-first, brand agnostic multimodal skin assistant. It can provide personalised responses to any skincare-related query, in real-time, 24/7. This is a game-changer in the industry, as we know from working with consumers that there are so many questions people have about each product, how to use them, combine them, the ingredients etc.
Both technologies are licensable for brands and retailers, to offer their customers this high quality level of personalisation.
How did your collaboration with dermatologists like Dr. Justine Kluk shape the development and accuracy of the AI?
To build accurate AI, you firstly need high quality data, and then you need subject matter experts to help with the training. When thinking about skin health, no-one knows more than a consultant dermatologist, so it was essential for us to harness this knowledge to build into our technologies.
Dr Justine Kluk was instrumental in the development of our AI skin analysis system. We worked together to understand how to recognise and differentiate different skin conditions, as well as how these conditions presented on different skin tones, and at different severity levels. This formed the knowledge base from which our AI was trained, to ensure the highest accuracy.
What challenges did you face in developing AI capable of providing ingredient-level skincare recommendations?
One of the challenges was collecting the data on what ingredient combinations were truly working for different types of skin profiles. Measuring real skin experiences over time is the most reliable and comprehensive way of doing this, but took several years to amass a dataset of this significance. Our data shows that 94% of people who used Renude’s service saw visible improvements in their skin, which is a reflection on the quality of recommendations being made, which is the same knowledge base that fuels our AI recommendations.
When working with B2B partners, we needed to build a system that could analyse any new ingredient list and extract key properties to assess suitability. As a formulation chemist, this was the really fun part for me! There are a lot of ingredients used in skincare formulation, so it was a large scale project to build the analysis system in the way we did, really leaning into the properties of ingredients. This is one of things that really sets our approach apart from other technologies in the market.
The ‘shelfie’ analysis feature is a world-first — how does it enhance the customer experience and build trust with brands?
As we have built a significant level of ingredient and formulation understanding into our system, we pride ourselves on being able to analyse an ingredient list to extract key properties of a formulation. This includes the texture, active ingredients, free from attributes, and overall suitability.
Using our Computer Vision capabilities, consumers are able to upload a photo of a product in their routine and get feedback on how well suited that is to their skin, and what they should be combining this with to achieve the best results for their skin.
We’re excited to be bringing this industry-first innovation to market to meet the evolving needs of the modern consumer. 
How do you ensure the AI can assess different skin types and tones accurately to provide inclusive recommendations?
There are two aspects to this; 1) training AI with the foundational knowledge on what skin concerns look like on different skin tones, and 2) showing the AI enough examples of these skin concerns on different skin tones.
Bias in AI is created when there are gaps or biases in the data used to train it, or gaps in the knowledge. There have been a lot of examples of this that have made the headlines, but in reality if you know what biases exist in data before you develop AI, you can mitigate these.
When thinking about skin tone, there are data gaps that exist in the real world, including in medical training, for example how redness presents on dark skin. To avoid building these real world biases into AI, they need to be mitigated up front before you start training your models. This was one of the areas we worked with Dr Justine on to ensure we had clear visual guidelines on all the skin concerns that we detect for, at different levels of severity across all skin tones. This meant we were training the AI from the start to work for everyone. As you share more examples, the accuracy of the algorithms continue to improve as long as they have that high quality foundational training.
What impact have you seen so far from brands using Renude’s technology, and how do you see this transforming skincare in the next 5–10 years?
We license our technology to brands and retailers, to provide AI skin support via e-commerce websites as well as in-store and for events.
What we see is that end consumers are really interested in engaging with AI, and curious about the outputs. We have baked a lot of ingredient knowledge into our technologies, which can be seen throughout the customer journey. This builds trust in the recommendations we make, leading to higher rates of conversion, and larger basket sizes.
Some of the results we’ve achieved for B2B customers include a 2.5X increase in Average Order Value, 40% uplift in Average unit price, 3.5X increase conversion, and an average of 5-8X ROI for the technology overall.
Our AI Skin Chat has been designed to set a new standard of personalised customer care for the beauty industry. While customer care teams do a great job, consumers have an increasingly short attention span so are looking for answers in real-time, otherwise they’re at risk of leaving the site and never converting. On top of this, the knowledge of an individual is limited, particularly in the specialism of skin, where there are many different disciplines – e.g. dermatology, aesthetics, formulation – which are involved in providing the highest quality advice.
We expect to see conversational AI technologies becoming the market standard for customer care, and are glad to be leading the way as the only specialist conversational AI tool for the beauty industry.
What’s next for Renude? Are there any upcoming innovations or developments we should be excited about?
Our AI Skin Chat has had an upgrade and the new version launches next month. We have already previewed this to key brands and retailers with incredible feedback, so we’re excited to take this to the wider market.
We’re also exploring new categories but can’t say too much more on that yet…
As a female founder in the tech and beauty space, what advice would you give to other women looking to innovate in this industry?
I love the beauty industry, I’m sure everyone that works in the sector will agree that people are (on the whole) extremely helpful, passionate and collaborative. We’re so lucky to have lots of incredible champions of our mission, who are willing to shout our name in rooms we aren’t in.
For anyone new to the beauty industry, I’d say build your network and don’t be afraid to ask for help.
