I do have some ideas, but I need some advices.

Current situation and roadmap
Avalanche has been a work in progress for almost nine years now. This is the first "game" we publish. For those of you not familiar with it, Avalanche is pure fluff : a multi-systems, multi-plots, non linear, calendar based campaign. It is almost an illustrated novel, but not really. It more looks like a story-board : a high level, with no details plots and setting. This seems well appropriated for rpg. So, by definition, as it is for storyboards, Avalanche is an unfinished product (and presented as such) : the "finished" version is the game at your table. Will Avalanche ever be finished as a product ? No.

So, where are we ? Avalanche is built to be incorporated into a website orientated toward a service to manage a campaign. Bad news : there is currently no roadmap for the website. Good news : we've got enough stuff to sell it without the service, as a pdf. We're planning to release this pdf somewhere this summer in its beta version (see below for the strategy). What will this include ? Here we go :
- Informative pages on how to use Avalanche (different aspects of the product, impact at the table, possible games, interaction with systems, introduction to the world ...). Those are not written "officially" yet. But most of it has been produced in a way or another on forums (mostly the forge and story-games), my blog and as part of my teaser.
- Example of play (AP post in fact). I've already got a bunch of those. Mostly need editing.
- The first part of the story (Avalanche has four), including the essence (what is usually called the setting). I'm talking here about what is translated and ready to be released. We're talking about 120K word here.
- Various maps (geographic, geo-strategic, rmap).
- Around 350 illustrations out of the 450 needed to cover this first part. The essence is all illustrated (a hundred big illustrations).

Note two things :
- There are no systems whatsoever at the moment. We're planning to work on this and adapt TSOY for Avalanche, but it is not planned for alpha and beta.
- There are some hundred illustrations missing. We're not doing them for alpha and beta.

We're aiming at a very high quality product.

Alpha
First question would be why do alpha ? But let me first address why not do alpha :
- As Avalanche is "novel like", getting "your story stinks" will not change much things. It is kind of too late (that said it is not what we get so far from the readers).
- Also getting "you should add robots or lizardmen, that would be cool" won't help much either. We're not adding robots nor lizardmen.

So, why do alpha :
- In formative pages are still to be completed. I believe that supporting players might really help me to know what to throw in there and how.
- AP would be more than welcomed.
- Testing in the context of a given system is always different and needed in my publishing model.
- Maybe we could see emerge a community around it. Just maybe.
- I don't like the term buzz, but alpha seems like a right time to do it as we can see the release of the product coming.

Of course, alpha would be in contact with some kinds of ashcan : no layout, not much of informative pages (apart from my blog) but fully illustrated.
Do they seem like good reasons ? Does the timing seems right ? Do I miss something ?

Now, here's a couple of questions about those alpha :
- I would be looking at a very few numbers of alpha testers ; we have the number five in mind. Does this seem enough ?
- The thing I really hate about alpha is getting people knocking at your door, looking at the product and not even playing it. So, I thought of dealing with alpha somehow like I deal with illustrators/master mapper/translator : through a contract. The clauses would include : no diffusion of the product, engagement to play and report games through AP with a certain amount of words (we're thinking about a mandatory three sessions), permission for me to use the AP in the product, credit for the alpha, beta and final version for the free for the alpha. Is this a usual practice ? How should I approach this ? Would I frightened people doing it ? Is a contract "overkill" ?
- Whom should I target for this alpha ? As I see it, I have, more or less, two choices : propose it to various communities (something like here at the forge) which would give me a larger range for systems (and maybe problems) but less support to the players (they would be less able to help each other) or propose it to an already existing community of players (for instance BW or d20), but thus "putting all my eggs into the same bags" (I'm not sure this translates well from french). What's your recommendations ?
- Where should I manage the support and AP report ? Setting out my own forum for this doesn't seem like a good idea. We plan on doing it, but more in beta phase. Any thoughts on this ? What worked for you ?

Beta
I'll present this since it is somehow related to the alpha.
Beta should come somewhere this summer and would be considered to be our first release. Beta would include informative pages and layout. But not the missing hundred illustrations.
What we want to do is sell the beta, limited to some hundred copies. And that, for three main reasons :
- Finance the final illustrations.
- If we cannot sell a hundred copies of the beta, that's very bad news. This also allow us to test our marketing strategy with strict, and limited, objectives.
- Encourage the beta to play it. We believe that people getting out money out of their pocket are more incline to invest time in the product.

As you can notice there's some kind of "bounty" model here (the bounty is the final version).

Finishing the beta, before opening for the final product, we should be able :
- To pay for and complete the final hundred illustrations.
- To pay for and complete the translation of at least some parts of the continuing story. Enough to go into periodical mode.
- Adjust some informative pages and see what kind of support are needed.

The beta would have access to the final version in pdf for free when it is released. We might even throw in the possibility to buy the paper version at "the cost".

Is fixing such objectives a good idea ? Is this common practice ? Does the deal seems fair ? Does this encourage play ? Does presenting it as a bounty seems a good strategy ?

Opening the final version
When the beta is completed (objectives reached), we close down the selling, prepare the final version (which should be a work in progress), distribute it for free to the alpha and beta and "re open" selling.
Now, and let's take a step at a time, a whole new horizon opens to us and we can "deploy" our publishing strategy, based on selling a periodical and throwing in some CC aspects. But that will be the purpose of another thread.