4 ms·

Big companies seem to step in the same puddle, every time they try to set up one of these "breakthrough incubator" projects. They pick fascinating projects. The

by GCA10 10y ago

Big companies seem to step in the same puddle, every time they try to set up one of these "breakthrough incubator" projects. They pick fascinating projects. They hire lots of smart people and give them great autonomy. But they inadvertently set up incentives that lead to the repeated creation of LINS (lavishly impractical non-solutions). And then the plug gets pulled. I think the key problem is that the research team starts optimizing for periodic demo days with the boss. That's a sheltered environment in which clever (and easy to demonstrate) technical features are rewarded, and real-world annoyances involving customers, social norms, regulators, pricing, etc. are put off for "later." Not only are ecosystem problems not addressed, they aren't even vigorously considered. The technology behind Google Glass was quite clever and some variant of that idea may eventually work. But what arrived on the market was LINS to an extreme. You can find similar failures in Detroit's concept cars, Paul Allen's first go at Vulcan, Xerox Parc's grossly overpriced STAR, etc. I'm not sure how to fix this. Within the big-company budgeting system, it takes a daring CEO to allow super-innovative projects to break the usual rules about desired five-year ROIs. Once such a CEO takes a stand, it's really hard for him/her to get out of the way. Suppose these CEOs do miraculously set up a system that approximates the scrappy, minimum-viable experiments of a true startup, with the marketplace being the true boss. That still is problematic. When a big company is pushing MVPs into the marketplace, the public scrutiny and scorn makes it really hard to recover from an awkward start -- and keep iterating in peace.

jackcosgrove 10y ago
"I'm not sure how to fix this." I'd say it's an unsolvable problem, and that's good! What breakthrough projects are good at is pushing the envelope and inventing whole new technologies. These technologies are then experimented with by smaller firms a few years later, and the winners of this scramble bring the technology to market. This is a spaghetti against the wall kind of market research; I think it's impossible to solve the problem of how to identify a new market and win it over. You can definitely have an informed approach that reduces risk, but it's not a technical problem that can be systematically solved. There will always be risk where there is reward. If it was possible to systematically identify new business opportunities, there would be many more players in the entrepreneurship business, and commensurately lower profits. In fact, an "informed approach" can sometimes lead you to discount real opportunities as too pie-in-the-sky. We need different risk propositions to make an economy with more niches for different kinds of people. I don't think the difficulty of identifying new breakthrough businesses (presumably with breakthrough profits) is a problem at all.
htrp 10y ago
Secret investment (or VC it out). Reputational risk is a real thing. I do wonder how much coverage glass would have gotten had it not had the Google halo around it.
bunderbunder 10y ago
I don't think it's a particularly wise idea for big companies to try to emulate startups or startup incubators. A key thing about startups is that they don't have much riding on them. If they fail, you can just dissolve the company, and the worst thing that happens is that people lose money and people lose jobs. If a big corporation tries to do the same thing, there are a lot more encumbrances that elevate the risk levels quite a bit. There's a mothership that wants to protect its reputation, there are internal social pressures to start throwing money at problems, there is risk of tort situations that originated in the "startup" having wide-ranging repercussions throughout the company. No sane company is going to look at that minefield and conclude that it really is safe to sit back and let people do the "move fast and break things" thing without having to worry about meddling suits. The risk/reward for shareholders just isn't there.
baccheion 10y ago
Brilliant internals? Maybe (never checked). Worthless product that solves nothing? Definitely. Developing great systems or internals, then trying to find a use for them is one thing. Google Glass was a crap product without a purpose (from the beginning). That aside, Google doesn't know what to do with clever, ingenious, or well-Engineered systems/tools/libraries/components/etc. They'll wrap them in sh-t, then use them to try to promote it. Even the way they try to promote their products is laughable. It sounds like poorly done attempts to spin absolute crap to make it seem as though it's something else. Why don't you just have what you're pushing not be absolute crap? Can't do that? Don't know how? I fail to see what was innovative or groundbreaking about Google Glass and the other crap they pitch as being such. Driverless cars? Is the driving of a car the real problem, or even a problem? Driverless cars sound cool in a way (would grab my attention), but the technology has long been created. What's on display now is a lack of ability to think about or come up with ways to take such technology and put it to use. What they're also making clear is that they're investing not because they see the many potential future applications, but because they don't know and are just picking something to try to say "me too," or to have its failure clearly demonstrate the pointlessness of trying to innovate, or because this or that name is associated with it, etc. They have no real idea what they are doing. They just do stuff. Also, the moonshots are strange, because while they like making themselves seem progressive or innovative, they don't do anything without a "business reason."
- adwn 10y ago
  > Driverless cars? Is the driving of a car the real problem, or even a problem? Really?! You don't see the advantages of a self-driving car? What about taking a nap, reading, or watching a movie while you drive to work? What about having a couple of drinks and getting home safely? What about being dropped off in the middle of the city right at the front door, and your car looks for a parking spot, maybe 10 blocks away? What about the potential for vastly reduced risk of accidents? What about cars that communicate with each other to improve merging, thereby avoiding congestion? You really don't see the value in this? A self-driving car that handles traffic situations at least as well as a human is the holy grail of personal transport. You couldn't have picked a worse example. > Driverless cars [...] the technology has long been created. No, that is completely wrong. If it weren't, driverless cars would already fill the streets.
toast0 10y ago
Google could allow a small group to work on something with use of Google's full software stack, and HR etc, but host it in google cloud (or proxy through google cloud so it looks like it's in google cloud) and don't mention it's from google (and segregate the user data away from google) If it gets traction, you can reveal it's a google project or fold the learnings into appropriate google projects or spin it off (although spin off is more complex if you did build it on top of the usual google stack) Allow something like ten people years and a reasonable amount of hardware, and when you run out of that, come back for more budget or make revenue. I haven't worked at google, but from experience at large companies, having the basics set for huge scale helps a lot (and avoids choice paralysis) so you can concentrate on doing the thing you wanted to do.
- tln 10y ago
  I was going to suggest something similar. Provide a way for employees to work part time or on a sabbatical, while on their own company. The incubator could take a small %age like YC, and Google would get to keep employees that would leave otherwise, closer to home.
  - QuercusMax 10y ago
    That's pretty much exactly what "Area 120" is. Not sure how many projects they have in there right now, but that is precisely what the idea is.
morgante 10y ago
This is exactly what Alphabetization is meant to solve. A federation of companies can more effectively allow individual divisions to operate like startups. Additionally, it's a lot easier to push experiments when they're not attached to the Google brand. Personally, I think Google should just funnel all the profits into GV.
- GCA10 10y ago
  Ah, but my point was that they don't operate like startups. They may adopt a lot of startups' superficial trappings, but in their first three years, they aren't sharpening their objective by being cash-constrained and exposed to the evolving tastes and distastes of the outside ecosystem. They're semi-generously funded while building whatever will amuse Sergey and Larry. That's creates LINS problems, again and again.
  - morgante 10y ago
    I agree that they don't operate like startups enough, but it is in fact the goal of Alphabetization. You're right that it's not working well enough. That's why I think they'd be better off simply making GV the largest private equity/venture capital firm in the world.
wilde 10y ago
Yeah. It seems like the real solution is to set up an investment fund and maintain distance from the incubator.

by GCA10 10y ago

jackcosgrove 10y ago

"I'm not sure how to fix this." I'd say it's an unsolvable problem, and that's good! What breakthrough projects are good at is pushing the envelope and inventing whole new technologies. These technologies are then experimented with by smaller firms a few years later, and the winners of this scramble bring the technology to market. This is a spaghetti against the wall kind of market research; I think it's impossible to solve the problem of how to identify a new market and win it over. You can definitely have an informed approach that reduces risk, but it's not a technical problem that can be systematically solved. There will always be risk where there is reward. If it was possible to systematically identify new business opportunities, there would be many more players in the entrepreneurship business, and commensurately lower profits. In fact, an "informed approach" can sometimes lead you to discount real opportunities as too pie-in-the-sky. We need different risk propositions to make an economy with more niches for different kinds of people. I don't think the difficulty of identifying new breakthrough businesses (presumably with breakthrough profits) is a problem at all.

htrp 10y ago

Secret investment (or VC it out). Reputational risk is a real thing. I do wonder how much coverage glass would have gotten had it not had the Google halo around it.

bunderbunder 10y ago

I don't think it's a particularly wise idea for big companies to try to emulate startups or startup incubators. A key thing about startups is that they don't have much riding on them. If they fail, you can just dissolve the company, and the worst thing that happens is that people lose money and people lose jobs. If a big corporation tries to do the same thing, there are a lot more encumbrances that elevate the risk levels quite a bit. There's a mothership that wants to protect its reputation, there are internal social pressures to start throwing money at problems, there is risk of tort situations that originated in the "startup" having wide-ranging repercussions throughout the company. No sane company is going to look at that minefield and conclude that it really is safe to sit back and let people do the "move fast and break things" thing without having to worry about meddling suits. The risk/reward for shareholders just isn't there.

baccheion 10y ago

Brilliant internals? Maybe (never checked). Worthless product that solves nothing? Definitely. Developing great systems or internals, then trying to find a use for them is one thing. Google Glass was a crap product without a purpose (from the beginning). That aside, Google doesn't know what to do with clever, ingenious, or well-Engineered systems/tools/libraries/components/etc. They'll wrap them in sh-t, then use them to try to promote it. Even the way they try to promote their products is laughable. It sounds like poorly done attempts to spin absolute crap to make it seem as though it's something else. Why don't you just have what you're pushing not be absolute crap? Can't do that? Don't know how? I fail to see what was innovative or groundbreaking about Google Glass and the other crap they pitch as being such. Driverless cars? Is the driving of a car the real problem, or even a problem? Driverless cars sound cool in a way (would grab my attention), but the technology has long been created. What's on display now is a lack of ability to think about or come up with ways to take such technology and put it to use. What they're also making clear is that they're investing not because they see the many potential future applications, but because they don't know and are just picking something to try to say "me too," or to have its failure clearly demonstrate the pointlessness of trying to innovate, or because this or that name is associated with it, etc. They have no real idea what they are doing. They just do stuff. Also, the moonshots are strange, because while they like making themselves seem progressive or innovative, they don't do anything without a "business reason."

adwn 10y ago
> Driverless cars? Is the driving of a car the real problem, or even a problem? Really?! You don't see the advantages of a self-driving car? What about taking a nap, reading, or watching a movie while you drive to work? What about having a couple of drinks and getting home safely? What about being dropped off in the middle of the city right at the front door, and your car looks for a parking spot, maybe 10 blocks away? What about the potential for vastly reduced risk of accidents? What about cars that communicate with each other to improve merging, thereby avoiding congestion? You really don't see the value in this? A self-driving car that handles traffic situations at least as well as a human is the holy grail of personal transport. You couldn't have picked a worse example. > Driverless cars [...] the technology has long been created. No, that is completely wrong. If it weren't, driverless cars would already fill the streets.

toast0 10y ago

Google could allow a small group to work on something with use of Google's full software stack, and HR etc, but host it in google cloud (or proxy through google cloud so it looks like it's in google cloud) and don't mention it's from google (and segregate the user data away from google) If it gets traction, you can reveal it's a google project or fold the learnings into appropriate google projects or spin it off (although spin off is more complex if you did build it on top of the usual google stack) Allow something like ten people years and a reasonable amount of hardware, and when you run out of that, come back for more budget or make revenue. I haven't worked at google, but from experience at large companies, having the basics set for huge scale helps a lot (and avoids choice paralysis) so you can concentrate on doing the thing you wanted to do.

tln 10y ago
I was going to suggest something similar. Provide a way for employees to work part time or on a sabbatical, while on their own company. The incubator could take a small %age like YC, and Google would get to keep employees that would leave otherwise, closer to home.
- QuercusMax 10y ago
  That's pretty much exactly what "Area 120" is. Not sure how many projects they have in there right now, but that is precisely what the idea is.

morgante 10y ago

This is exactly what Alphabetization is meant to solve. A federation of companies can more effectively allow individual divisions to operate like startups. Additionally, it's a lot easier to push experiments when they're not attached to the Google brand. Personally, I think Google should just funnel all the profits into GV.

GCA10 10y ago
Ah, but my point was that they don't operate like startups. They may adopt a lot of startups' superficial trappings, but in their first three years, they aren't sharpening their objective by being cash-constrained and exposed to the evolving tastes and distastes of the outside ecosystem. They're semi-generously funded while building whatever will amuse Sergey and Larry. That's creates LINS problems, again and again.
- morgante 10y ago
  I agree that they don't operate like startups enough, but it is in fact the goal of Alphabetization. You're right that it's not working well enough. That's why I think they'd be better off simply making GV the largest private equity/venture capital firm in the world.

wilde 10y ago

Yeah. It seems like the real solution is to set up an investment fund and maintain distance from the incubator.