NØIE’s scientific and data-driven approach
NØIE is the result of science and knowledge on skincare that has been combined with machine learning algorithms and invaluable feedback from our community of users. Especially the closed loops of community feedback are what makes NØIE truly unique and what continues to improve the solutions for both our existing and new users.
This data-led approach has already been embraced in so many other industries as well as in our everyday lives. But for some reason it has never really taken off in the skincare industry. This means we have missed out on the opportunity to make important progress - until now, that is.
The result of our approach is that we now know exactly what ingredients to use for which skin issues. That is why we can make customised solutions for each of our users, and why 86% report an improvement in their skin from using NØIE.
But NØIE was not created over night. It has been a long journey throughout the past 6 years. Here’s how it all started.
Reviewing the market
Back in 2016 the two founders Chris Christiansen and Daniel Jensen set out to learn more about skin diseases and the existing solutions as a project within LEO Pharma’s Innovation Lab. Skin diseases are the 4th most common cause of all human diseases (Tizek et. al, 2019). Christiansen and Jensen began investigating by gathering 2,938 existing over-the-counter products aimed at alleviating various skin issues ranging from acne to psoriasis. Then began a 2-step screening process of those products.
First, the two founders along with Christian Riemer (who is now our Head of Research & Development) and their team identified and excluded the products containing potentially harmful ingredients such as perfumes, well known irritants and suspected endocrine disruptors specified by an external panel of experts. What remained were only 19% or 566 of the 2,938 products.
Second, they tested the perceived efficacy of the remaining 566 products. This was done by making the products available in a webshop and collecting data on the users after purchase regarding their demographics, symptoms and skin conditions, as well as asking for their feedback on the products after a 30-day trial period.
Testing the learnings
Analysing all the data points resulted in some valuable learnings about which products seemed to have an effect on the different skin conditions. Now was the time to test those learnings on a smaller group of people and the team initiated a pilot study with the aim of validating the findings from the data collection.
The research team invited 10 people from the UK suffering from either psoriasis or eczema to participate. They were offered individual solutions combining different over-the counter-products selected from the 566 mentioned above. The test period of 8 weeks involved a weekly update on the condition of the participants’ skin as well as an ongoing adjustment of the product recommendations.
After the test period, the severity of the participants’ skin conditions decreased and the requirement for medical treatments such as topical steroids dropped on average by 40%. While in itself it was an achievement that non-prescriptive solutions were effective when used correctly, it was also the starting point for the NØIE algorithm.
Building the NØIE algorithm
Individual skincare recommendations based on thorough data analysis seemed to have a visible effect, and so Christiansen and Jensen founded NØIE in 2019.
NØIE started to develop and produce its own formulations and offer its own personalised solutions to people with skin concerns based on the solid data foundation and extensive knowledge about the efficacy of specific ingredients and synergies among the ingredients.
In order to enter NØIE’s platform and receive a product recommendation, the user fills out a detailed online survey, taking all relevant clinical aspects of the skin into account. This data is then processed through machine learning algorithms, to assign personalised formulations for the individual. These formulations are then evaluated by users who give ongoing feedback on whether or not they experienced a positive effect on their skin. All of those data points provided the backbone for the NØIE algorithm.
Improving the results
The first success rate from the feedback was that 30% of the users experienced a positive effect. As the machine learning model continued and the feedback kept being fed back into the algorithm, this success rate on NØIE’s solutions leapt from 30% (2020) to 72% (2021) over the course of a year. Today, the success rate of NØIE's approach, using nothing but the right and unique combination of ingredients, is at 86%.
Since skin is such a personal issue, often with a psychological dimension, it is all the more important to improve the products and make informed recommendations to decrease the amount of time and money wasted as well as increase the overall success rate.
This in short is how NØIE used, and still uses, science, data, user feedback and machine learning to provide personalised skincare solutions.
Find a more comprehensive explanation of NØIE’s scientific foundation here.