I currently won't buy video content unless I know I will want to watch it again, or unless I have a reasonable expectation that multiple people will enjoy it once. That's because the current price point for most DVDs is too high for me. At $5 for a package with 4 different movies in it, I can take a chance on something out of the bargain bin, but $15-$25 for one movie is not reasonable unless I have already seen at least a portion of it and liked it.

So we have a problem. I won't buy something of unknown quality, and content producers don't want to show me something new unless I pay first. My dollars therefore don't go to them until after one of the following happens: I see it at the $2 second-run theater, I see it on television, I see someone else's licensed copy (incl. Redbox, Netflix, or public library), or I see a pirated copy. Nobody wants to buy a pig in a poke.

(Yes, that's right, the latter means that piracy can result in additional sales.)

Once I have assessed the quality of the work, I can decide how much I am then willing to pay to watch it again, on demand. If that's more than the current price of a disk copy, I might buy it. If not, there's only so long I am willing to wait for the price to drop, before my disposable income is already disposed of, and my available library storage space is filled. I might save it to a gift wish list, but it's more likely I'll simply forget it exists.

And it's not lost sales that kills you in a creative business, and it's not even people hating your stuff, it's people not knowing your work exists. So in my view, the burden is upon the content distributors to get my attention that first time, to signal a level of quality that will make me feel comfortable spending my money.

So here's the model I think will work:

- Establish a basic service for streaming permanently licensed content. For this, charge the marginal cost of hosting and outgoing bandwidth. Allow customers to suspend and reactivate their accounts at will, without hassle. Allow customers to download those titles and their licenses to local storage whenever they want. Make sure your application makes local versus streamed content seamless, so customers won't be tempted to use a different player.

- Allow customers to pay an additional $X to get $X in non-refundable service credit. For each purchase of service credit, give the customer N discovery tokens. Allow purchase of additional tokens with store credit. For example, $10 might get you $10 of store credit and 40 tokens, and spending $5 of that credit might give 20 more tokens.

- Allow customers to subscribe by automatically repeating a specific purchase package on a schedule they determine. Subscribers get bonus rentals and juicier discounts to encourage that option over pay-as-you-go.

- Customers may spend discovery tokens to temporarily license and stream content. One token should probably be roughly equal to one entire feature-length movie, or three hours of a television show. While temporarily licensed, purchase of a permanent license is discounted by $X/N for each token used.

- Allow some content to be streamed without tokens, and auction off a number of permanent licenses by allowing customers to secretly bid for them with their store credit.

- Use purchase/non-purchase and auction bid/price information to fuel the recommendations engine.

The bet here is that customers will never be content to just download their permanent catalog and use your application as a media player. They will always be paying you to discover new content, and paying more for you to recommend better reasons for them to pay you. During the time when someone is in maintenance mode, you are still putting your brand in their face and cultivating goodwill every time they watch. You are also offering a value-add over Netflix in that customers build up a sort of loyalty equity in your service, so that they build up a larger personal catalog as they continue to use it.