Freemium vs Ads? Or both?

I am developing an online utility application. My partners and I have struggled to come up with a profitable business model. We were thinking that we could sell server space to allow people to put their files they make with our application there, but then I realized that there is a slight problem. What if people use our application, but download their files to their desktop? My vision for my website was that the application would have the same features all throughout the pricing system, but that you would need to pay to use more than 500 MB of server space. But again, what if people use 500 MB of server space, and then download all their files, and then start over again? How can I turn this around, but still allow users to utilize all of the same features? Also, to calculate profit margins, what percentage of users do you think would probably use the free service, and what percentage would not? I looked up Skype's business model, which I knew to Freemium as well, and they have 35% of users using Skype for business, and 65% of users using Skype for personal use (according to this site near the bottom of the article http://voiceontheweb.biz/2009/03/skype-business-model-revealed-at-ebay-analyst-event/ ). Does a 35%-65% freemium model sound right to you? Or would it be a different percentage because it's on the Internet, and not a software that you download? If you think that going freemium is a terrible idea, then what would you suggest instead?

Actually, I'm a huge fan of Freemium. A lot of people do it wrong though.

First, you're not going to have a clue what percentage of users will use the free vs the premium version of your service until you start. Once you have it up and running, you can tweak the features of the different plans to fit yours and your customer's needs. If you don't want to screw over the old users (and you don't, trust me), just grandfather them in.

Personally, I would look at Wufoo.com's model for Freemium. The execute it brilliantly. I personally started as a free service user, and have since upgraded to their tier 3 experience, "Bona Fide".

Having at least 3 service plans as opposed to 2 will grant you a lot of flexibility for revenue as well as for your users, which is what we want when we're on the interwebs.

Now, for the big mistake most people make that I talked about earlier.

When implementing Freemium, you need to make sure that you don't give away the farm with your free version. There should be great features withheld, or I should only be able to use it so many times per month etc. HOWEVER, the product still needs to be functional and get the basic job done. Don't give it all up, but don't make it suck so bad I won't want to use it. The only job of the free version is to get me hooked on your product.

Thanks Matt!

If you are a self-funded startup, I would be very wary of a pure freemium revenue model: you don't have the cash to burn to survive as you scale so that the 1-2% who pay provide significant revenue. 35% is way, way over the top: for example Evernote started at 1%.

Advertising (or the lack thereof) can work, but again, depends of building a huge number of regular visitor and the storage/advertising approach has been done to death.

Put ads on the free product and remove them on the paid version. That way you get revenue whether your customer is paying (from the customer) or not paying (from the advertiser).