[petewailes]

There's something important in the latter half of this: understanding that this isn't useful as a guide, but as a warning.

No-one realistically sets out to create "a Google". People set out to solve a problem. Sometimes that problem turns out to fuel vast business, sometimes not. You can guess at the size, but sometimes you're going to be wrong (hello Twitter).

You can aim to replace a business, but it's near impossible to do it because you know what the future of that thing will be (hello Windows Mobile). Even ARM didn't know smartphones were going to be a thing, at the scale they are. ARM ate Intel's lunch, not by going after smartphones, but by making processors designed to work with virtually no power. Smartphones just happened to come along and need that.

You can't intend to be disruptive. You can aim to become a big business, but that's out-competing, not disrupting, and they're not the same (hello Snapchat). The market tells you you were disruptive after the fact.

[rhaps0dy]

Good comment. However for future readers, I wish to point out that "ARM are Intel's lunch" is probably instead "ARM ate Intel's lunch".

That might save some confusion :)

[petewailes]

Edited. Good spot!

[mrec]

Also, is there a "don't" missing in "you know what the future of that thing will be"?

[petra]

But still, you can try to steer towards being disruptive when your choose your idea: you select unserved/underserved markets, or markets you can serve at lower profits than the incumbent, and do so using technologies you think can improve enough in performance/quality to compete with the incumbent while still having lower costs(which is a good way to choose a tech anyway).

And sure it doesn't guarrantee anything, but it increases your chances of disruption.

And with that, you still need to build a strategy and find barriers to entry. because (a)companies are more prepared towards disruptive innovation today , and (b)if you don't have good barriers to entry, once you become a threat, the incumbent will copy you.

[logicallee]

>No-one realistically sets out to create "a Google"

but lots of people unrealistically set out to create "a Google" and some of them succeed. Who never would have if they had had a realistic goal.

my goals are completely unrealistic. most people with such goals fail completely.

[petewailes]

Even if you succeed, that's anecdotal. Telling others to do the same wouldn't be good advice, in the same way as a lottery winner telling others to play wouldn't.

[logicallee]

If one in ten thousand people who set out to create a billion dollars of value for society and themselves succeed, and the others fail but use the skills they learned to lead gainfully employed, happy lives, why is that not good advice? Society nets a billion dollars, and the expected return is let's say 1b/10k = $100k, though the distribution of outcomes is skewed. So why not? What's the objection?

[wslh]

I think this is changing and a company like Google can see what we call "the future", based on virtually infinite data points. I am not saying that these data points are all possible futures but they give a clear advantage over entrepreneurs with minimal data. True, you can't predict the real future but Intel could act earlier with more information. Andy Grove wrote about this on his "paranoid book".

[stcredzero]

Google can see what we call "the future", based on virtually infinite data points.

If Google can manage them usefully, those points are clearly not virtually infinite. (Otherwise, you're diluting the meaning of the word.)