Peter Kämpf is wrong, He correctly edited his answer.
I just finished my masters degree, writing about Servitization, and used Rolls-Royce PLC as a "textbook" example.
In my thesis, I included an interview with Rolls-Royce Service Researcher Alistair Forbes.
He explained the basis model as such:
Some people misunderstand our business, they think we rent engines out to customers. We don’t. When an airline buys an aircraft they buy the whole aeroplane. That includes the engine. But they pay a fee to us for every hour they fly the aircraft, for every hour the engine is flying, they pay a set fee. We well translate all the expected costs that we would have, that might be on a contract [...] that might be we expect, okay, an engine doesn’t need an overhaul – unlike like you service a car. You don’t actually drive it that much. Airplanes fly around 14 hours a day, and needs an overhaul every 5 to 7 years. Its expensive when it happens but they all stay reliable.
It’s partly this change in business model, before it was 3-4 years, now we try make a more reliable product, time between overhauls increased 5 to 7.
Now if I’ve got a 10 year contract with a customer, and I say, that you pay me 200 dollar for every hour you use the engine. And nothing ever happens, I’ve got 200 dollars per pair and done nothing. But supplied a fantastically reliable engine. The customer is happy and we’re happy. if I make an unreliable engine, the customer is unhappy and they say the engine needs an overhaul every year, and I don’t incline to overhaul it once or twice. That’s bad for both of us.
For reference, the economist wrote an article regarding how RR managed to stay afloat, here they also mention speculations regarding selling engines at a lose, to capture marketshare. http://www.economist.com/node/12887368
(Approx. half way through the article)