A curious leak on Friday about the ‘ChatGPT for Doctors’ has reasonable people wondering what dilution even is. OpenEvidence, an AI chatbot for doctors, has reportedly received offers to fund the company with $200M at a $20B valuation.
The company is unlikely to proceed, in part because a round would dilute founders and other shareholders.
The quick math on this is, of course, selling $200M in shares at a $20B valuation corresponds to 1% dilution. If existing investors are lukewarm on the company, then existing investors would use the dilution charade to sell some of their shares.
We could expect that this press conversation was either on behalf of the company, existing investors, or prospective investors. For the company, the press would read that it’s hanging out a shingle to get more (and higher) offers. For prospective investors, the press could be to push the process forward, create some doubt for the company if other offers don’t materialize, or create some chatter on X about the absurdity of the dilution excuse—complaining about 1% dilution.1
For existing investors, placing such a story in the press could catalyze more bidders that reach out to existing investors—which, given the positioning, would then be about cashing out existing investors. Prospective and existing investors would get to talking and—given the dilution constraints—realize that the only way for the new investor to own part of OpenEvidence without the company taking dilution would be to buy shares from existing investors.
The founder cares so much about dilution? Well, investors can pitch the company on an anti-dilution round. The new investor buys from an existing investor (no dilution) and some of the existing investors’ shares also return to the company, where they can be retired (reverse dilution). All of a sudden, all other shareholders now own a greater percentage of the company—and like magic, the problem is solved.
Anthropic is hiring doctors now. Abridge and others are circling around this space. There may be a fantastic business here advertising drugs to doctors while they work, but we don’t need to pretend that this business doesn’t have any risks. OpenEvidence needs to license content from medical journals for the chatbot to be effective and the journals might wake up to the fact that they can charge quite a bit more for the content they are providing. Are there cheaper alternatives? Maybe you could get Bryan Johnson to license data to you, but you’ve kinda lost the plot at that point.
The fundraising signal carries across other parts of the business, particularly potential advertisers. Drug companies are an interesting potential investor for OpenEvidence with the posture of being stingy on dilution. A drug company might realize that it could secure some advantages in exclusivity of the advertising channel for certain verticals in exchange for a bigger revenue commitment that OpenEvidence could flip to investors in boosting the next funding round’s valuation. Why not have all the cardiologists always see your company’s drugs? Seems like that could be compelling enough for a drug company to pre-pay and own the channel.
Some assumptions are pretty heady already:
OpenEvidence is currently selling less than 5% of its ad inventory, suggesting it could generate billions of dollars in annualized revenue if it sold the rest, the person said.
Yes, exactly. Public companies like Reddit, Pinterest, Snap, Roblox, and NextDoor all have a similar dynamic. In theory, could all those public companies sell more ads? Yes. But at a certain point the ads that can be sold are sold as much as possible. Maybe less than 5% of ad inventory is also just a way of saying we’re not selling many ads at all and we don’t have the sales motion to do it or we are too worried about doctors leaving if we bombard them with ads. If the ads work, the advertisers plow as much money as they can through the channel. There is no compelling reason for OpenEvidence to gate it if it can deliver a great experience for doctors and advertisers.
It’s probably much more important to be thinking about how to avoid diluting the company’s strength than about diluting the cap table in a way you’d need a stethoscope to pick up on.
-
The reporting claims the founder owns 58% of the business, so there is no risk that this funding moves the founder below a majority level of control. ↩