I have a thesis that the success of startups that operate in a market with strong network effects is often more about luck than skill and that, therefore, repeat success on the part of the founder(s) is by no means guaranteed.

It's like when a first-time golf player scores a hole-in-one. If you try to get them to repeat the feat, they can't because they don't actually know what they did to get the ball in the hole - they just whacked the ball in the general direction of the tee.

Now, take a thousand ball-whackers (startups) all targetting the same tee (market). If the network effects are strong enough, one of them is going to win simply by getting their ball closest to the pin. They don't have to actually get it in the hole - as long as it's the closest, the network effects will take over and push the ball into the hole.

Result: The ball-whacker is perceived as a genius and VCs line up to invest in his next startup.

However, if the first success was, indeed, down to luck, there's no guarantee that the second attempt will be equally (or, indeed, anywhere near as) successful.

This isn't to say that all first-time hole-in-one ball-whackers are just lucky. Some are clearly naturals. The trick is distinguishing between the naturals and those who were simply lucky.