When people say “Go big or go home” they’re rarely actually thinking they’ll have to go home. If they actually saw that as a real risk, they’d probably go at least slightly smaller.

I’ve long said that the growth of Groupon — and to a lesser degree LivingSocial — were testing something much broader than whether daily deals were a sustainable category. They were a Harvard Business School case study waiting to happen on the merits of insane, rapid, expensive, and aggressive international expansion in the Web 2.0 era.

Groupon made LivingSocial — and everyone else in comparison– look conservative. The company aggressively poured hundreds of millions of dollars into International expansion and while they excelled in some areas, other efforts were plagued with scams. Others still were textbook cases of how not to enter a new market. The big shock when Groupon filed its S1 was how badly the international divisions were performing and how much money they were losing. That lead to several management shakeups, but seemingly no real answers.

Back when everyone was calling Groupon the fastest growing company ever, you could see why the No. 2 player LivingSocial felt compelled to try to keep up. But its 10 percent staff cuts today, show that it too is under pressure to perform. While there were only a few dozen cuts outside the US, that overspending has to be coming back to haunt Living Social as it struggles with profitability.

Are startups drawing any lessons from these flameouts? Or just blaming the international failure on the lameness of the daily deal space and moving along?

I hope it’s the former. International markets hold a ton of promise, and many of the entrepreneurs to be found there should be taken seriously as competitive threats. But the execution risks aren’t to be taken lightly. You rarely hear a startup say, “We failed because we were just too focused on our core business.”

Granted, it’s hard to know exactly what lessons to draw when it comes to international expansion. Nearly every approach to capture the global Web audience has failed in some way.

Pretty much everyone whiffed on how to play international in the Web 1.0 era. Most consumer Web companies waited until they were dominant in the West and public, and then turned their eyes to the rest of the world. They felt they had all the time in the world. The downside was they were slow moving bureaucratic beasts by this point and loathe to give local markets they autonomy they needed to thrive.

Ironically, Yahoo did better than anyone else, particularly in China — the hardest market of them all. Rather than try to build out their own Yahoo or hire a local team and take away all their authority, Yahoo simply bought a healthy stake in Alibaba. The value of that stake made up a sizable chunk of struggling Yahoo’s market cap until recently.

In the Web 2.0 era there was significantly more pressure to go wide, fast. Not only did startups hope to do a better job than the first generation of Web companies, but the attack of the international clones was out of control. Thanks to a proliferation of international capital, huge online populations around the world, and — once again — dramatically lower costs of starting a company this time around, no Internet company felt it had the luxury to simply build in the US and wait.

Many still waited a comparably conservative long time. Facebook was in that camp until it was pretty solidly dominant in the US, and still doesn’t have an answer for China. It wrote off Orkut’s popularity in Brazil until it was ready to throw down. When it did, it handily beat then long-in-the-tooth Orkut.

Twitter, too, has been fairly conservative, given how broad its international usage has been since the beginning.

Interestingly, it’s the commerce related companies that seem to be so much more aggressive in recent years. Amazon only recently entered some major markets like India. Netflix is barely available anywhere outside the US. But this time around, Fab, Uber, and HotelTonight and other relatively young companies have been rapidly expanding overseas.

Granted, a lot of this is limited to Western Europe. Still, the latter two are particularly surprising because each market requires its own local network effects to take off, and neither are close to penetrated in the US. We’ve heard, too, that Square founder Jack Dorsey plans on spending the bulk of his time overseas next year, working on Square’s global expansion. Given the remaining execution risks in the US– and the fact that Square is priced for perfection valuation-wise– that might be a risky gamble.

Part of the reason these companies are doing this is because they feel pressure. One reason investors argue that multi-billion dollar Web valuations haven’t been out of control is that there are billions of people online today. The simple fact that America has some 300 million people total tells you that companies have to expand if they want to justify that logic. Meanwhile, using iOS or Facebook’s platform for distribution drags you international whether you are really ready for it or not.

But another reason is they feel they can go international more easily with global platforms like Facebook and iOS and payment providers like Braintree and Stripe also going international so quickly.

But just because you can do something, doesn’t mean you should. Unless your business has very strong network effects, the fear of a local clone being first to market and gaining a foothold that can’t be broken later shouldn’t outweigh every other problem a startup CEO faces. The once fever-pitched fear of an attack by the Samwer Brothers has chilled now that they’ve closed their Fab clone and their Airbnb attempt has mostly failed. Those thousands of Groupon clones in China are now gone or on the ropes as well. It’s easy to argue, in hindsight, Groupon should have focused on the US longer. As it stands, they’re left with a bunch of mediocre businesses here and overseas.

In the US, the best entrepreneurs will say they keep their heads down and focus on what they are doing, not what competitors are doing. (Foursquare’s Dennis Crowley had a great, must-see articulation of this at our October PandoMonthly.) Why should that be different in a market you don’t know that has a much higher risk of failure?

My gut tells me the pendulum is going to swing back, and some CEOs are going to start to rethink this next year. Particularly if growth capital becomes more scarce or more expensive, valuation-wise.

[Image via xdr.com]