Update: We Raised $8M From Our Crowdfunding Campaign And Grew Active Subscribers 150%
This is a follow up story for SelfDecode. If you're interested in reading how they got started, published over 2 years ago, check it out here.
Hello again! Remind us who you are and what business you started.
Hi, my name is Joe Cohen and I’m the founder and CEO of SelfDecode. The company was born out of the journey I went through to decode myself.
SelfDecode provides users with an integrated AI platform that provides personalized health recommendations based on a combination of your DNA, labs, and environmental factors. In addition to providing a direct-to-consumer service, SelfDecode has signed numerous business contracts licensing its technology to a variety of health companies.
I started SelfDecode in 2016 with a mission to make precision medicine standard in the healthcare industry and now we’re providing people and practitioners with all the tools you need to understand risk factors and build a personalized action plan for optimal wellness.
Users are provided with comprehensive genetic reports about their health predispositions, lab test analysis, and tracking tools, and lifestyle assessments to get the fullest picture of their health. Based on these results, SelfDecode provides personalized diet, supplement, and lifestyle recommendations.
This is in major contrast to the traditional one-size-fits-all approach to health in traditional medicine, in which healthcare strategies are developed for the average person, with much less consideration for our unique characteristics and risk factors.
In 2021, we launched a crowdfunding campaign and raised a total of $8M and in 2022, our revenue remained stable at about $2.6M, as we used our extra funds for R&D and building out our tech, which I’ll explain more below.
Tell us about what you’ve been up to. Has the business been growing?
Since 2021, we’ve grown our active subscribers by more than 150%, and currently have 18,000+ individual members and 500+ health professionals.
In 2022, SelfDecode averaged a 13:1 return on every dollar spent on marketing.
We’ve kept our marketing team and efforts fairly small over the past two years so that we were able to utilize the additional funds from revenue and our previous raise to improve our core technology, which allowed us to create the best genetic analysis models currently available.
Before our launch of SelfDecode version 2.0, cutting edge polygenic risk scoring by looking at millions of variants was simply missing. Companies were using prediction methods only slightly better than a coin toss.
Below, you can see the results from the tests that compare our models (Orchestra and Selphi) to the leading industry standards.
It's important to note that most direct to consumer companies are not even using industry standards or legitimate methods.
Fig.1: Orchestra vs. state of the art LAI methods. (a) Percent recall and precision for ancestry deconvolution by FLARE (navy), Gnomix (light blue), RFmix (green), and Orchestra (red) across 6 generations of admixture. Each generation is represented with a star shape. The number of points on the star corresponds to the number of generations (0 - 6). Orchestra outperforms other methods in the 1000 Genomes Project (1KGP) dataset with 16 populations (left) and our larger custom data panel with 35 populations (right). Accuracy (%) per population for the 16 populations in the 1KGP dataset (b) and the 35 populations in the larger custom dataset (c). Populations are ordered by mean accuracy across all methods (cross). The overall accuracy for each reference panel is shown on the right.
As seen above, Orchestra markedly outperforms other leading ancestry methods in both non-admixed (generation 0) and admixed samples (generations 1-6). It has ~15% better overall recall and ~14% better overall precision than the next-best model. It also retains high accuracy across all tested populations, with a remarkable ability to distinguish between closely related ancestries.
Orchestra achieves an accuracy of over 75% for 100% of populations within the 1KGP dataset and 75% of populations within the custom data panel. The other models struggle with about a third of the populations, where their accuracy falls below 50%.
Fig. 2: Accuracy of Selphi vs. state of the art Imputation methods. Imputation errors across chromosomes 1-22 for Beagle 5.4 (blue), Impute 5 (magenta), Minimac 4 (Yellow), and Selphi (Green) for different minor allele frequencies (MAFs) (a) and in different superpopulations (b). Target and reference samples are obtained from the 1000 Genomes Project (1KGP) 30x reference panel. Errors are shown as deviations from the average number of errors across all 4 methods. Selphi outperforms all other methods for all minor allele frequencies and all four super populations. Selphi is in particular better at imputing rare variants (MAF < 1%).
Whole‐genome sequencing fully captures genetic variation but remains expensive and commercially unviable for larger datasets. Instead, direct to consumer companies generally rely on cost-effective genotyping arrays which capture a sparser set of variants. Genotype imputation then fills in the gaps, by inferring missing variants from a relevant reference panel. The accuracy of imputation, therefore, is crucial for all downstream applications.
As seen above, Selfie outperforms leading imputation methods across all allele frequencies, performing exceedingly well for rare variants, where all other models struggle. This is of importance because rare variants are more likely to be of medical significance. Selphi also performs better in all tested super populations: Africans, East and South Asians, and Europeans.
Proper ancestry and imputation are important pieces to getting accurate genetic predictions, and companies don't use these in their algorithms.
They typically will look at a few variants to predict something, but these methods are outdated and scientifically flawed.
Additionally, we made other significant advancements in our product and technology, as we:
- Optimized our AI and machine learning technology to provide clinical-grade Polygenic Risk Scores
- Combined genetics with conventional labs and self-assessments within SelfDecode health reports
- Implemented Ancestry-Informed Polygenic Risk Scoring Models
- Improve our genetic file analysis to impute over 83 million SNPs with 99.7% accuracy, outperforming academic and commercial models by 22%
It's important to emphasize that we focused on developing the corer technology because we felt this was the bottleneck in getting legitimate genetics into the hands of the consumer and the healthcare market.
Because of our advancements in our technology and valuable consumer-ready software, SelfDecode has massive B2B potential. Already in 2022, we were able to secure 25+ B2B partnerships with an estimated value of $69M over 5 years.
We secured these partnerships primarily through outbound marketing efforts, using cold outreach to relevant businesses.
In addition, we plan to get FDA approval for various tests so that they can be incorporated into mainstream medicine.
Throughout the year, we also worked on improving the product, dialing in our subscription offers, and introducing targeted and relevant upsell to our customers, nearly doubling the average lifetime value from $247 to $447.
Some changes we introduced to increase lifetime value include:
- A more premium subscription plan that offers more 1:1 support from a practitioners
- Offering lab tests for sale, at a discounted price for our current customers
- The ability for our customers to order personalized supplements based on their DNA results
Since 2021, we’ve grown our active subscribers by more than 150%, and currently have 18,000+ individual members and 500+ health professionals.
What have been your biggest challenges in the last year?
Educating consumers on the difference between legitimate, scientific, and quality genetics products vs illegitimate products.
Most people think that when you buy something, maybe it's 10% better than something else, depending on the brand.
In the genetics field, it's the difference between going to a palm reader and getting a legitimate test that has actual value.
Almost all companies in the direct to consumer health market are using illegitimate methods, and none are using the latest/most cutting-edge methods. This is why direct to consumer genetics gets a bad reputation.
Most consumers don't understand what is legitimate and what isn't in the field, and most companies capitalize on this by building a product with shitty science, and doing their main investing in marketing.
How does a consumer know what is legitimate and what isn't? Even doctors and scientists who don't have a specialty in population genomics don't know.
It's also not possible to figure out what is legit before a consumer buys.
So, educating consumers is a tough process with an inherently new and complex product.
What have been your biggest lessons learned in the last year?
There are very talented, and sometimes more affordable, people out there who would love to work with you. Don’t count them out just because they can’t come to work in your office.
This year, I’ve learned the importance of me staying involved in promoting brand awareness for my company. While building out our tech, I became very involved in hiring, HR, and product development. I took a step back from being involved in social media, conferences, and podcast.
However, in 2022, I jumped back in and we started noticing a significant increase in traffic and sales from organic and brand awareness channels, including social media. I began sharing my personal health story again, along with the strategies I used to achieve optimal health, and got great feedback from our audience.
I even re-launched a new podcast that I’m hosting, The Joe Cohen Show, and have made great connections with industry leaders that have resulted in increased affiliate sales as our guests have started to promote SelfDecode.
What’s in the plans for the upcoming year, and the next 5 years?
As of Q4 2022, we have signed 25+ partners to license our technology out to other companies in the personalized health space.
Including:
- Direct to consumer genetics companies
- Direct to consumer lab testing companies
- Supplement companies that want to integrate precision health
- Health apps or devices that want to integrate precision health
Throughout the next several years, we plan to continue building out our B2B and D2C products, ultimately working towards larger health studies and clinical trials. This year, we will be obtaining patents and releasing several more scientific papers. You can see the research we’ve already released on our Media Page.
B2B Subscription Product for integrative Health Practitioners
- Client management system (2023)
- Integrative client reports & health recommendations (2023)
D2C Product Expansion
- Comprehensive health reports that combine multiple PRS into one report (current)
- Expansion of our recommendations feature (2023)
- Precision Health Forum - connects patients with similar conditions, genetics, and labs (2023)
Long term goals (2023-2027)
- Utilize our large user database to dramatically improve predictions and recommendations
- Conduct a health economic study
- Partner with corporations to reduce healthcare costs via preventative treatments.
- Provide a clinical decision support system to medical providers - tools that allow doctors to make better decisions
- Conduct clinical trials for FDA approval for patented diagnostic tests
What’s the best thing you read in the last year?
To be honest, I don't read books because of how busy I am.
I found that the best thing that I can do for myself is not gain knowledge, it's to improve how my body functions.
By improving how my body works, I have more energy, focus, creativity, and calmness - and am not distracted by getting sick or ill health.
In the past year, our technology got good enough that I was able to spend some time focusing on my health and using the SelfDecode technology to improve all aspects of my life.
Advice for other entrepreneurs who might be struggling to grow their business?
My best advice is almost always the same…hire the right people. When you hire well and surround yourself with a good team, you are almost guaranteed to see success.
I operate a fully remote team with employees all over the world and one reason for that is that I don’t limit myself to people in one city. I can select the best applicants from all over the world. There are very talented, and sometimes more affordable, people out there who would love to work with you. Don’t count them out just because they can’t come to work in your office.
Are you looking to hire for certain positions right now?
This year, we will be looking to hire a Marketing Partnership & Communications Specialist to help grow our affiliate program and secure more partnerships that promote SelfDecode.
You can read more about the position and apply here. Plus, you can always keep an eye on our Careers page for other open positions.
Where can we go to learn more?
If you have any questions or comments, drop a comment below!
- 4,818 founder case studies
- Access to our founder directory
- Live events, courses and recordings
- 8,628 business ideas
- $1M in software savings
- 4,818 founder case studies
- Access to our founder directory
- Live events, courses and recordings
- 8,628 business ideas
- $1M in software savings
- 4,818 founder case studies
- Access to our founder directory
- Live events, courses and recordings
- 8,628 business ideas
- $1M in software savings
- 4,818 founder case studies
- Access to our founder directory
- Live events, courses and recordings
- 8,628 business ideas
- $1M in software savings
- 4,818 founder case studies
- Access to our founder directory
- Live events, courses and recordings
- 8,628 business ideas
- $1M in software savings
- 4,818 founder case studies
- Access to our founder directory
- Live events, courses and recordings
- 8,628 business ideas
- $1M in software savings
- 4,818 founder case studies
- Access to our founder directory
- Live events, courses and recordings
- 8,628 business ideas
- $1M in software savings
- 4,818 founder case studies
- Access to our founder directory
- Live events, courses and recordings
- 8,628 business ideas
- $1M in software savings
