Sensible Groupon offerings don't have to make money, or even fail to lose money. They just have to do so within the bounds of a marketing budget, and provide a better ROI than what the budget would have gone to otherwise.
The amount of money a rational organization should spend on negative ROI activities is exactly zero. It's just as stupid to spend your money on a -25% ROI as it is on a -50% ROI, you'd be better off putting the money in a GIC.
So if you set aside a 10k yearly advertising budget, offering deals which make you lose 2k is perfectly reasonable - if you think the exposure and new customers are worth it. The advertising budget is already accounted in the yearly ROI, so you don't have to worry on each deal's bottom line being in red.
For the service industry in particular, hourly rates can vary drastically by geography and type of business, whereas for spas, restaurants, or other store-front businesses, it's clear whether or not you're getting an actual deal/coupon because you can compare the deal to the base menu or pricing packages.
Advertising costs money so whenever something is "advertised" you're going to pay more unless the advertising generates enough business to get an economy of scale greater than the cost of advertising. Advertising makes a lot of sense when you have something with high capital costs but low per unit costs. (Think CPUs or software based businesses). When you're buying a service such as cleaning where most of the costs come from labour it's going to be difficult to generate an economy of scale great enough for advertising to pay for itself. In businesses with out great economies of scale the increased cost of advertising is going to be passed to the consumer.
You're always going to get a better rate on a service business by doing a little leg work yourself and cutting out the "middle man".