Sourcing Undiscovered Talent
Find great people before the market prices them, then earn the intro with real work.
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 7
- Confidence
- 86%
The mechanism converts recruiting from a bidding contest into a discovery process. The first input is mission framing: state the largest honest version of what you are doing, because the best people are aware of their own potential and are drawn to meaningful work. The second is timing: lay out that mission before the space is crowded, so you can reach great engineers before the market has repriced them. The third is the sourcing loop itself. Instead of scanning the obvious field, you look one step adjacent, where people tinker on odd projects for their own reasons. You then spend real time understanding one person's work, think about it, and come back with a modification, a contribution, or a considered question. The output is a warm conversation with someone nobody else has found, and a candidate who responds because they were seen, not solicited.
Origin
Extracted from the Naval podcast episode Curate People, drawing on a July 2025 tweet that the job of a startup is to find undiscovered talent and distill it into a product, with the mission-framing pattern read off Elon Musk and Sam Altman.
Core principles
- 01If you can identify talent from afar easily, so can everybody else.
- 02A mission framed at its largest possible scope is itself a recruiting instrument.
- 03Being early to a mission is how you reach great people before the market does.
- 04Makers have taste in other makers, which is why sourcing cannot be handed to HR.
- 05Genuine interest outperforms outreach engineered to recruit.
How to run it
- 1
Frame the mission at maximum scope
State the ambition in its largest form rather than the nearest deliverable. Not going into space but going to Mars, not building chatbots but building AGI.
Pro tip Never waver from the framing in public; consistency is what makes it credible to the people you want.
- 2
Be early to the space
Lay out the mission before the field is crowded, so the best people can join while it still looks impossible to everyone else.
Watch out If the space is already crowded, skip to creative sourcing rather than competing on offers you cannot win.
- 3
Hunt one step adjacent
Avoid the obvious centre of the field. Look at the odd adjacent problem where people tinker for their own reasons, not the one everyone is already scanning.
Pro tip Pick an adjacency your own taste actually covers, so you can tell good work from noise.
- 4
Do the work before the outreach
Spend a day or two genuinely reading their code or paper, think about it, and return with a tweak, a piece of code they might incorporate, or a considered question.
Pro tip A good question earns a reply because the person is tinkering unnoticed and someone finally saw it.
Watch out Outreach that is transparently a recruiting sequence gets treated as spam by exactly the people you want.
- 5
Hand off to the closer
Once the relationship is genuine, bring in whoever in the founding team is the strongest recruiter to make the actual case.
- 6
Build a standing surface for builders
Create a place or ritual that puts you around founders and makers continuously, and do it because you enjoy it rather than as a recruiting funnel.
Pro tip Doing it for its own sake makes it authentic, which is what makes it work.
- 7
Expect and absorb a low hit rate
Accept that most approaches fail. The pipeline is justified by the rare person nobody else could have reached.
In the wild
Rather than scanning for people training better AI models, which is the obvious field, a co-founder looks one step sideways at someone applying unusual machine learning algorithms to micro weather forecasting. He spends a day or two reading their GitHub and paper, thinks about it, then emails with a small piece of code they might incorporate and a considered question. The person responds well because they have been quietly tinkering and someone finally noticed and engaged seriously. He is not doing this to recruit; it is what he does for fun, which is precisely why it lands.
→ A relationship with a strong, completely undiscovered builder that no recruiter would have surfaced.
An assistant is hired after being met at a restaurant, with no prior tech company experience at all. What was visible was that everything the person touched was quality and stylish and that they cared about it. The company took the chance on the basis of demonstrated taste rather than a matching resume.
→ A hire that no conventional sourcing channel would have produced, made on evidence of care and quality.
Common mistakes
Recruiting people who are already famous
By the time someone is well known on social media or fully credentialed, everyone is competing for them and the terms are set by the market, not by you.
Delegating sourcing to recruiters or HR
They cannot bring you weird people because the risk is too high for them and they lack the maker taste required to recognise unusual quality.
Faking genuine interest
Outreach engineered purely as a recruiting funnel reads as such. The approach works because the interest in the person's work is real.
Is it for you?
Best for
Founders competing for scarce technical talent who cannot outbid larger companies and need to reach people before they are legible to the market.
Not ideal for
Roles that genuinely require credentialed, licensed, or heavily regulated expertise where discovery-based sourcing adds no advantage.
From the transcript
“If you can identify the talent from afar easily, so can everybody else.”
“By the time someone's famous on Twitter, it's too late to recruit them.”
“So to be a great recruiter, you have to first be a great sourcer.”
From the episode
Curate People