As the chart of the day shows, the revenue growth is great but you don't want your net income to start looking like a mirror image. That could be the next Amazon (which was unprofitable for a long time, remember) or the next Webvan.
More concerning is how their last couple rounds of fundraising seem to have crossed the line entirely into private stock placements, paying off insiders and early investors.
April 2010 series F: $120m to existing stockholders, $15m of new equity
December 2010 series G: $810m to existing stockholders, $136m of new equity
February 2011, $17m to existing stockholders, $200,000 of new equity.
($70m of the December round went to NEA --- who have funded Tintri --- to redeem stock they bought in the series C and D offerings.)
Now, I wouldn't turn down the opportunity to lock in a cool billion of appreciated value, either, but still... if you can attract that quantity of late-stage money, why not use it to get to profitability? Or, if your IPO is on the horizon, why cash out $800m in December? (I suspect some of the series G investors are planning on a quick turnaround from a hugely-hyped IPO...)
Also dubious is inventing a new accounting metric so that you can claim the $240m you spent on advertising last year, and the $180m the first quarter, is just a temporary thing and doesn't reflect your true costs. It may or may not be the case that Groupon can keep its existing subscriber base happy at low cost.
Interesting risk mentioned in the S1: their position is that unredeemed Groupons aren't gift cards, and so are not subject to state laws about unredeemed balances. But, it's possible they might have to perform the required recordkeeping and increase their liabilities if they can't get away with that stance.