That Silicon Valley Founder Reminds Me of His VC!

Warning: Opting to fund one’s look-alike leads to lower returns

Silicon Valley has long been characterized as a “tech bros funding tech bros” ecosystem. About 4 out of 5 venture capital partners are men, and roughly 80% of the deals they fund go to startups founded by men. While this insularity is often ascribed to shared demographics — matching on gender, ethnicity or educational pedigree — a paper forthcoming in Strategy Science by UCLA Anderson’s Jane Wu shows the bias can also be driven by something far more literal.

Using facial recognition technology to analyze more than 93,000 potential deals between VC funders and startup founders, Wu documents a powerful preference for a familiar face. In early-stage investing — where hard data on a startup’s prospects is scarce — a significant (as defined in the paper) facial similarity between investor and founder boosts the odds of a startup getting funded by about 10%.

Looking beyond the obvious DEI implications, Wu observes that VC investors may pay a price for this behavior. Her analysis shows that every 10-percentage-point increase in the share of investors who closely resemble a founder was associated with a 14% drop in the odds of a successful exit — such as an IPO or a major acquisition.

“The pattern of ‘bros funding bros’ may come at a real economic cost.”
—Jane Wu

Wu leveraged data from Y Combinator, the Silicon Valley “boot camp” for startups. Twice a year, YC selects a batch of promising companies for seed money and intensive mentorship, culminating in a Demo Day where founders pitch a room full of deep-pocketed venture capital investors.

Because YC only accepts about 1% of applicants, every founder in the room has already been vetted for quality — making Wu’s findings, if anything, conservative. Getting to pitch at Demo Day means you have already scaled a high hurdle, yet the facial bias still shows up as a significant factor in who gets funded. 

Wu analyzed 12 YC batches between 2014 and 2019, along with the VCs on record as having been in attendance at each Demo Day, giving her nearly 100,000 potential deals to examine.

To measure how much these individuals looked alike, Wu collected professional headshots from online sources — YC’s directory, VC firm websites, LinkedIn profiles — and ran them through ArcFace, a state-of-the-art facial recognition tool. The software encodes each face into more than 500 distinct measurements: the slope of a jawline, the exact distance between eyes, the geometry of cheekbones. This produces a numerical “face distance” between any two people — a precise, continuous measure of how visually similar they appear, capturing nuances that broad demographic categories simply cannot.

‘Handsome Devil’

Wu then mapped the facial distance parameters for funders and founders. The preference for facial similarity remained a significant predictor of funding even after Wu applied controls for shared geography, gender, ethnicity and the industry of the startup. Wu also filtered out superficial photo similarities — whether both people wore glasses, had their mouths open or appeared to be the same age. The preference wasn’t about a shared “vibe” or photo quality; it was about the underlying structure of the face.

The effect was sharpest precisely when investors have the least to go on. Among early-stage startups — those with no product milestones, customer traction or revenue figures to share — the face distance effect was strong and statistically significant. Among more advanced growth-stage startups, where founders could point to concrete evidence of progress, the effect was weaker and imprecise. Facial similarity matters most when hard information is scarce.

To further explore the role of facial similarity in tech funding, Wu ran a lab experiment in which nearly 400 participants recruited on Prolific were given the role of VC funder. Each was shown photos of fictional entrepreneurs accompanied by a brief startup description. Some participants received “soft” pitches, in which the entrepreneur briefly described an idea they wanted to get off the ground; others received “hard” pitches, framed as startups already launched and in need of funding. 

‘I’m a Sucker for a Pretty Face’

Wu used the same facial-recognition technology to map the face distance for the participants and the fictional entrepreneurs.  With no conversation, no networking and no shared history to cloud the picture, participants were still more likely to say they would invest in entrepreneurs whose facial features resembled their own. The pattern held even after controlling for shared age, ethnicity and gender, suggesting that the visual connection is a distinct force of its own, independent of a shared résumé or a common social circle.

The familiar face bias was also stronger among participants who received the soft pitch than among those who received the hard one. This directly mirrors Wu’s finding from the YC data: The less concrete information available, the more a familiar face moves the needle. While not the direct focus of the lab study, she also found that participants viewed familiar faces as exuding more warmth and being more attractive, not that they were more intelligent or competent.

Wu’s research suggests that a VC model built on the value of personal relationships and gut instinct has a measurable blind spot. Investors may not consciously be seeking a mirror image, but the data suggests their judgment is being nudged in that direction, most powerfully when there isn’t a dashboard of performance to rely on. 

She suggests correctives such as blind pitch decks stripped of names and photos or awareness training to help investors recognize their own familiarity bias. Given the financial cost that the research documents, making that case to investors might not be a hard sell.

But the implications extend beyond investor returns. Venture capital’s gatekeepers remain overwhelmingly male and disproportionately white, and that imbalance has long shaped who gains access to early-stage capital. Past research, as Wu notes, shows that minority entrepreneurs often have strong ideas and ambition but struggle to break into the funding circles. Her research suggests a subtle, subconscious bias may also be operating at a more granular level, adding to the struggle minority entrepreneurs already face.