𝟮𝟯𝗮𝗻𝗱𝗺𝗲 𝗴𝗲𝘁𝘀 𝘀𝗼𝗹𝗱: 𝗘𝗻𝗱 𝗼𝗳 𝗮𝗻 𝗘𝗿𝗮, 𝗕𝗲𝗴𝗶𝗻𝗻𝗶𝗻𝗴 𝗼𝗳 𝗮 𝗡𝗲𝘄 𝗢𝗻𝗲 𝗶𝗻 𝗚𝗲𝗻𝗼𝗺𝗶𝗰𝘀 The sale of 23andMe’s assets to Regeneron for $256 million marks the closing of a defining chapter in consumer genomics and the start of a new era centered on integrated, multimodal healthcare data. 23andMe was once the face of accessible genetics. With over 15 million customers, it transformed DNA testing from a clinical procedure into a conversation starter at dinner parties. That kind of reach felt revolutionary at the time. And yet, in 2025, that immense dataset sold for just ~$17 per record. To put that in perspective, consider Tempus AI. Tempus is currently generating around $1,000 per patient per year in revenue from its data business alone. The key difference is multi-modal data: genomic, clinical, imaging, pathology, and beyond. This enormous disparity in per-record value ($17 𝘷𝘴. $1,000) reflects more than just different business models. It’s a sign of a fundamental shift in the biotech ecosystem: value now lies in data that’s both deep and connected, not just broad and isolated. We’ve seen this pattern before. Back in 2012, Amgen acquired deCODE Genetics for $415 million. Like 23andMe, deCODE was a trailblazer, using large-scale genotyping of Iceland’s population to draw genotype-phenotype correlations. That acquisition, over a decade ago, valued deCODE’s data at a significantly higher per-record level than what 23andMe fetched in 2025—despite 23andMe having nearly 20 times the customer base. Why? In 2012, deCODE and 23andMe were at the frontier of human genetics. Genotyping was the dominant tool for discovery, and large sample sizes meant power and potential. But by 2024, genotyping alone had become table stakes. The field has moved on (whole exome and whole genome sequencing, e.g. GeneDx) toward contextualized data: electronic health records, pathology slides, radiology images, social determinants, and treatment outcomes. In short: the era of single-modality discovery is over. Interestingly, there were signs of real competition in the bidding process. 23andMe's co-founder and longtime CEO made a last-ditch bid for $74.7 million. Regeneron’s $256 million winning bid suggests there were other bidders at the table. What were the motivations for their bids? Would’ve been great for the peanut gallery if the winning bid had come from Theranos 2.0, with Billy Evans and Elizabeth Holmes. Seems like we are closing the book on the age of consumer genotyping and entering a chapter defined by multi-modal, clinically-anchored, AI-ready data. The value will go to those who can unify messy, disparate data into something useful at the point of care or in drug development. It’s not the end of genomics. It’s just the end of 'omics based solely on nucleotide sequences. https://lnkd.in/gexqZPKV
Peter Farago
6mo
I was close to getting a 23 and Me report many times but never ended up doing so. Should I be sad or relieved?
Paul T.
6mo
Gautam Kollu - great insights and prescient Hidetoshi Watanabe Marc Iyeki
Very interesting!
Larbi Gallagher
6mo
Not bad
Andy Cosgrove
6mo
Reductionism science is evolving... the era of single-modality discovery is certainly changing.
Sue Ann Molero
6mo
thanks for the post Gautam Kollu and sharing your insights... interesting times -- but not sure I agree about Theranos 2.0 -- maybe better we close that chapter for good?!
Tony Gordon
6mo
Great commentary Gautam. Data without clinical context/demographics/phenotypic info becomes just data and this is exactly 23andMe’s issue.
Ashok Gopinath
6mo
This is a great insight Gautam, one that will find it’s way as the next revenue stream but more importantly a great value proposition in healthcare.
Great perspective, Gautam. So, do you think the juice is fully squeezed on single modality genomic data?