The Seed (Venture Investing) Rules

The following figure self-summarizes venture guy Vinod Khosla’s investment returns by amount invested, and then stratified by whether he had a board seat. The upshot: His highest returns came disproportionately from investments where he put in less than $1m, and from where he had a board seat.

seed vs series a investing

The latter isn’t much of a surprise, as the relationship between board seats and investment performance are well know, but the former is interesting. Too many venture investors don’t want to play in initial rounds where they can’t put in the requisite $2-$6m, so they don’t even look at sub-$1m rounds. As Khosla’s figures show, returns from seed rounds are impressive, to say the least.

Caveats: First, the period included the latter years of the tech boom, so there is a huge potential confound here. Second, any venture investor who has success rates in excess of 60% across the board, and 100% in seed, is an anomaly by any reasonable definition of the word.


  1. Also makes it clear why he doesn’t need to be a partner in a VC fund any more.

  2. PK> Happy 4th!
    Where did you get the data ? I agree with you, the 100% success rate coupled to that cash on cash return is out of proportion.

  3. One or two major home runs in the seed financings could really skew these results. Wasn’t Vinod Khosla an early investor in Google? If Google is included in the

  4. Hey Jeff — Thanks. Came from a public PPT deck of his I had kicking around. I think it’s still available from the KPCB/Khosla Ventures site, so I’ll try to link it in.

  5. I agree that a smaller investment brings higher return – an investee will maximze use of investment proceeds then and try not to overspend :)

  6. I’d love to know how he defines “success.” I was on the receiving end of several of his pitches while he was building syndicates and can say, for a fact, that they were not all successful.

  7. My mistake, I read the slide wrong. I don’t know for certain that any of the deals I looked at were ones that he seeded with less than $1m.
    I’ve heard a quote, often attributed to Steve Jobs, that more companies have died from indigestion than starvation. I agree.

  8. oops, I mean David Packard, not Steve Jobs. This is starting to feel like a Monday!

  9. A theory: perhaps Khosla will ‘always’ re-invest up to at least $1 million, whenever a company is struggling. (In for a dime, in for a dollar.)
    Then there’d be a guarantee all non-successes move to the “>$1M” category.

  10. agree with Andrew Fife, one data point could skew those results quite a bit. 20-30 data points doesn’t really make for empirical observation… probably need more like 200-1000 to really start drawing any definitive conclusions.
    also, other commenter’s point also well taken — not clear whether this data includes only initial and/or later rounds… ideally we should be looking at first round invested.
    lastly, given the changing nature of private equity, returns from vintage 91-95, 95-99, 00-04, and 04-onward likely substantially different due to market conditins.
    vinod’s a smart & unique guy — his returns probably shouldn’t be used to infer typical market behavior.
    – dave mcclure

  11. Andy Freeman says:

    The relevant question is how profit per expense, which is likely to be f(time, money). It’s likely that small companies take about as much time as larger ones and Khosla only has 24 hours/day. That may force him to bigger deals.
    Unless, of course, he’s measuring his success by number of successes, not dollars.

  12. Actually what is even more interesting is how much in total was invested and see if that correlates with return.