Correlate for acne
Acne self-experimentation app
Myself and 3 data scientists
1+ years (ongoing)
Founder, Team Lead, Researcher, Designer, Developer
Correlate is my passion project which came as an outgrowth of having many friends suffering from chronic diseases (e.g. Crohn's, IBS, depression, anxiety, acne). After an initial round of research and several rounds of self-experimentation, I decided to first focus on acne.
People with acne want to understand and fix their skin. However, they lack the power to:
- understand how their daily decision affect their condition
- share useful and convincing observations with their healthcare providers, and
- contribute their expertise to the broader knowledge community.
Doctors want to cure acne. However they are incentivized to treat symptoms and not fix the internal problem, because
- medication is easier to comply with than behavior change
- prescribing medicine allows doctors to bill at higher rates, and
- they believe in evidence-based treatment and few studies have focused on non-pharmaceutical treatments of acne.
Our mission is to help individuals perform scientifically robust self-experiments in order to a) find what works for their skin and b) be able to legitimize their observations with doctors and the broader medical community.
Concept Refinement: The Beginning
In the beginning, I thought Correlate could be useful for multiple diseases and had a data collection and analysis approach that I thought could work. However, I had questions:
- Would people want this product?
- Would doctors be receptive?
- And most importantly: Would my data collection and analysis approach work?
Here are the methods I used to find answers.
To answer these questions, I interviewed numerous people with chronic health conditions. This gave me a solid basis to understand a) their attitudes toward western medicine, b) their methods for managing their condition, and c) how they track data and analyze data relating to their condition.
Would people want this product?
In the interviews, I found a subset of participants were already tracking data. They were interested in tools to help them understand that data better, which was a good sign. They were interested in the product, but with caveats. This meant I had more work to do.
Would doctors be receptive?
Doctors were generally supportive of the concept, but wanted to make sure it a) didn't interfere with the doctor-patient relationship and b) didn't create more extraneous work for themselves.
Would my data collection and analysis work?
In order to answer this question, I designed and performed numerous self-experiments (or n-of-1 studies) in collaboration with a data scientist. During this phase, we came to tentative conclusions but the effectiveness of the method was still unclear.
It was clear from my UX research and self-experimentation that there was promise, but the concept needed refinement before building.
Concept Refinement: The Middle
Narrow the scope
I determined it would be best to focus on a single disease, based on conversations with people in the industry and on the fact that my analysis method might need to be more disease-specific to be useful. It was at this point I decided to focus the app on the condition I had studied most: acne.
Improve testing methods
I modified my methodologies based on scientific literature and ran another 4 rounds of self-experimentation.
Update target user
Since my primary user was now an acne sufferer, I needed to narrow my user research to that audience. Acne.org proved to be a wonderful source of people who were willing to share their experiences with me. Interviews from this group formed the basis for updated personas.
Refinement 4: Refine by Testing
To see if my concept would work for other people, I created and refined the concept over time with target users. Here is some of the design work in that process.
The refinement of Correlate is on-going. It continues to be 💛 some of the most fulfilling work of my career. 💛
If you're interested in following or using Correlate, sign up for updates on the Correlate website.