Free trials aren't free to the company. It's a promise given to customers that they will be able to make use of company's assets without payment. Company normally has to pay for both the actual usage that occurs and the option of usage it has granted to it's customers so far. Trial subscription and expiry mechanism lets company control the maximum amount of free usage they have agreed to at any point in time.<p>Let's say you came up with a new and innovative service that let's people publish pictures they took with a mobile phone into the cloud. Because you calculate it will cost you on average one dollar per user account per month to provide the service you decide to charge your users $5 per month.<p>To attract new users you also decide to give them opportunity to try the service before they commit. You have a few options:<p>1. Let users upload the pictures in reduced quality for free unless they convert into premium. You're worried that this strategy will distort your user base: people who actually need best possible quality will not get a true impression of your service. On the other hand you will have to pay indefinitely for the storage and bandwidth of a whole lot of inferior quality pictures. Even worse the policy will attract whole group of non-paying users that will be content with the inferior picture quality and will make a heavy use of your service.<p>2. Let people upload up to 50 pictures free. In which case again you'll end up with having to pay for non-paying customers storing their pictures forever. Even worse some unscrupulous people will end up opening hundreds of accounts and some kid in Russia will write a utility working on top of your service seamlessly opening and amalgamating free accounts for a single user.<p>3. Set a free trial period of 30 days when users can upload pictures for free. After 30 days they will either have to convert into paying customers or download their data because you're going to automatically purge the free accounts. You're worried that 30 days may not be enough for some people to evaluate your service, but then you decide to look at the conversion stats; after-all if the prospect couldn't make up their mind after 30 days maybe the need wasn't that pressing and you shouldn't be wasting your company resource trying to convert them?