I think the burning question is, what do you do with private repositories that are created during the open beta? Will I have to pay when Codaset comes out of beta to access my code again?

Still to decide on that one. But one likely options is to allow you to keep any private projects created during the beta and not have to pay for them after the beta ends. But then any new private projects would be subject to charges.

How does that sound?

That sounds like I'm going to create about a hundred private repositories, just in case :) Maybe add a grace period in which you can either open-source your code or upgrade your plan?

Nice! ;)

A grace period is a nice idea.

I really like the idea of allowing us to keep our current projects. Depending on how much money I’m making, I may consider a paid plan, but it makes me feel really secure in using this service to know that I can probably expect to retain use of my private repos. Thanks for all the goodwill!

I think allowing one private repository after the grace period would be a big feature over vs. github and codebase. If you allow one repository you can be sure that a lot of developers would use it and come back more often to us it. While they do so they'll see all those other features you're going to implement and you have an advantage when they decide what to use for their company. You could limit the size of it if you fear that people use one for every stuff they have or only allow one account to work with it on a free account.

Further, be sure to read several chapters of "Predictably Irrational" by Dan Ariely ( isbn10: 0007256531 ) before making up the plans, I'm sure you'd be astonished about the information in there. It'll help you to find the prices and ranges of your plans very well.

I think that the models set by GitHub and HoptoadApp are solid bases. A typical company like mine expects to pay around 20-25 dollars per month for 20-30 projects.

I'm with Youssef on this one Joel, I think one free private repo would be a great way to set Codaset apart from the competition.

I would gladly pay $100/year for:
10 private repos
2 GB space
SSL

I think having at least one private repository in the free plan would be a major... selling point, for lack of a better term, over GitHub.

I agree with rspeicher, lots of people want to play around with a private repository before forking (hehe) out the cash for a paid account.

I think your prices should compete with ProjectLocker. ProjectLocker has the lowest prices for private repositories with a lot of space, and even though they're not the most feature-rich at the lower prices, that attracts a lot of small development firms.

Another issue to consider is that the size of the repository does not necessarily correlate to the number of developers. From my experience, there are some small teams that generate very large repositories, and vice versa. It may have to do with different policies regarding storing 3rd party binaries and outputs in the repositories. For example, a lot of .NET projects I've worked on stored the nunit, nhibernate and other references dll files with the repository to ensure version compatibility between all the team members.

I would like to see a price model based on size rather than number of repositories, if I have 20 small repositories that don't take much space why does it cost more?

It would also be nice to know what you are doing as far as backups, security, etc...

Agree on that the pricing model should be based on size rather than number of repos. A repo limit seems arbitrary while space makes sense. I currently have more than 20 repositories on another host, and I tend to create a new repository for every lite insignificant project I start. Paying for a super-duper-enterprise account doesn't really seem to match what I do.

So then if every one wants to pay for space and not projects, what sort of space are people currently using?

Jimstr has an excellent point. I like to have a lot of repositories, and if I paid for the number of ones I use, I'd be bankrupt. But I would be happy to pay for space that I use. I'd love a pricing model where I under a free plan, I can have maybe three private repositories with a total storage limit of 300mb or something. Then, paid plans would have unlimited private repositories with limited storage, graduating toward a more expensive totally unlimited plan.

I don't really know a lot about how you plan to run your business, so these suggestions may not apply very well. These are just the kind of things I'd be really happy to see… :)

These are all brilliant suggestions, and exactly the kind of feedback I need, thank you all.

I can have maybe three private repositories with a total storage limit of 300mb or something

And how much would you be willing to pay for that?

I don't really know. My response was poorly worded, but I think that that sort of thing would fall under a basic, dirt-cheap plan, like $3-$5/month. In the past, I've seen several SVN hosters that give this sort of thing away for free, but I've never seen a really nice, high quality Git hoster do that. So I think you could get away with charging something along those lines. After that price point, though, all limitations on the number of possibly repositories could be lifted, but quotas on space could be retained and gradually expanded.

From my point of view, most of this - # of repositories, # of private repositories, # of users, disk space - is trivia. I can (and have) set up my own git server and my own gitweb and get all that for the cost of hosting. But as someone running a growing consulting biz, I'm willing to pay for two things: (1) Not having to administer my own system (or pay for my sysadmin's hours to do it) and (2) As close to 100% uptime as humanly possible. Of course, this sort of premium business may not be common enough to be worth chasing, but I'd happily pay $50/month for something demonstrably more reliable than GitHub. Alternatively, there are ways you could make Codeset attractive as a "GitHub fallback" - though whether that's a marketing position you want to take I don't know.

I'd second the predictably irrational book, take a look at that and google for the ted talks on it - there's a lot to learn, and since you have something to apply it to, it will be more then a trivial bit of information.

Personally - I'd like to have one private repo for free. It would keep me using the service even when I didn't need it. And then I'd pay for space and more private repos (with client support - so I can share my code with my clients). Maybe make all the paid plans have unlimited repos and a space limit.

I would give nothing for free. The perception is that it is not worth much if you give it away. Provide better service than others and charge for it.

I'm with the folks that would rather charge based on space rather than number of repositories. I have close to 50 private repos I host myself. This is just because git makes it so simple to create a repository, so why wouldn't I? Collectively, these repos take up very little space.

An alternative I tried to float by the GitHub guys that I think would work, too, is to have a concept of "active" projects. So, maybe I host 50 repos on codaset, but I can only have 3 active ones and pushes are rejected for the non-active projects. This matches my use case considerably better.

I think that it's premature to determine what a "reasonable" price would be as there are too many unknown factors and other points for comparison with other services.

But I know what I can afford and I know what I need.

I can afford ProjectLocker's Seed "service level", which offers 2.5GB, unlimited repositories (both Git and SVN) and 5 users for 25 USD per year, trac integration and very cheap priority support options. That's YEAR, not month. It doesn't give me everything I need (automated builds, email tickets, better extensibility, reporting, etc.), but enough to get along.

For the same price points and Codaset's current service offering, projects would simply grow wings and flock on over en masse. I'd even pay a bit more for better extensibility (which seems to be one direction Codaset is actively pursuing).

GitHub, in comparison, is way too expensive for me and seems relatively unreliable.

Please never limit the number of repositories (private or public), for any price. Keep in mind that very often private repositories eventually turn public. Allowing infinite repositories encourages forking, increased communication between developers and more experimentation. I'd even suggest that repository counts are increasingly fictitious: very often you use multiple repositories for different "sub-products" and/or projects.

I'm with you, Noam. GitHub is really nice, but far too expensive. Premium, expensive services like Github and Basecamp are really awesome for small businesses that make a lot more money than I do. I feel like I would be happy to pay for something like Codaset, but I would never pay what Github charges. Not when I could pay A LOT less and host it all myself.

@Jonathan, @Noam, I don't agree with you. Codaset is a tool. If you use the tool, you should pay for it. When you say that you have 50 git repositories, what's in them? If it's anything of value, paying $15 a month to keep them safe would seem like a bargain.

May you can read my article on this subject: http://ariejan.net/2009/10/25/they-are-just-tools-people/ I'd like to hear what you think about it.

@Ariejan, your points are well taken. As for those 50 git repositories, I like to experiment a lot. I suppose if there were a limit on repositories, I would just consolidate, but that might be semantically counter-productive. Thanks for writing the insightful blog post; I've commented on it.

As a freelance webdesigner and working alone I keep my repositories local. But I would love to be able to put them online and a) access them anywhere and b) have a secure backup that way.
As I work for ridiculous low prices myself and my income is neither high nor constant I have often looked at these kind of sites, but always found them too expensive. Some have low prices or even free hosting, but then the repository space is way to small. Keeping finished projects in the repository becomes impossible and that is not acceptable since at any time a client can call me and ask me to make a quick change on their website, for which I need direct access to that repository, no matter how long ago I finished the project. I can give you these facts:

1) My number repositories grow over time. Most of these will be inactive mostly.
2) The avarage reporitory size is 20MB
3) I could pay a maximum of 5 euro per month for that on avarage.

Now, I am not saying I think that is a reasonable price, but it is all I can afford (for now).

What is the status on paid plans? We are looking at GitHub and Kenai unless you can give us an idea on pricing.

Thanks,
John

Sorry, but no decision has been made yet, but there will be one very soon. Please bear with me.

While I cannot vouch for it's profitability, you could attempt to have a user defined pricing structure.

I've seen it done with restaurants and other businesses where customers pay what they think the service is worth. So perhaps you bind the clients to paying x price per month, for a year at a time minimum. And since you may have credit card processing fees and so forth, you create a lower bound of a couple dollars to ensure you can cover that. Past that, you let the users decide what to pay you.

As for providing a free repository, it'd be useful, especially if you could have multiple collaborators with you, say 5 perhaps. But on the other hand, you would then have projects utilizing strictly

From what I've heard, it seems that often times the customers overpay what the business would charge. Naturally you could put some ads and some tracking information around the site, stating that an average users pays x for y and so forth, which would be interesting to track. Not to mention you'd most likely get some extreme publicity out of implementing a pricing scheme like this.

It's a bit of a radical idea and certainly is dependent on a belief in humanity of some sort but if you can stomach it, it might be worth a shot. Codaset is already proving to be worth using and it is clear that a lot of people are more than willing to pay for it.

A major downside would be that if you find yourself not being able to meet your operating costs +, you may be put in a situation where you must change the pricing scheme, which would lose you some customers in the transfer no doubt.

All in all it's a lot to weigh - which you clearly know. Best of luck deciding.

Regardless, a single free private repository for any user would be excellent, with perhaps a contributor limit of 5-6 people. That way a smaller project could really start itself here and then expand via payment, yet you wouldn't have gigantic private development teams making off with a service for free. Payment by space would be excellent. I'd certainly be curious to see some statistics based off of the beta what an "average" repository size, with and without respect to the contributor count.

With the payment plans I would certainly start no lower than 5-10 private contributors if feasible.

I agree with a limit on the number of private repos and contributors on a free plan.
In a paid plan, I agree with an unlimited repos and contributors with a pricing structure based on storage size.

You can also add nicer things on the paid plan that you could charge for users, though I still don't know what that is.

Well I can confirm that there will definately be a free basic plan, that will most likely let you create an unlimited number of open source projects. And it will definately let you create at least one private project.

More to come...

I would also like to add, please do not artificially limit the number of projects or repos on your paid plans.

Please take a look at http://repositoryhosting.com (however they use trac which I do not like)

Many single developers who have many small projects do not utilize much resources. Please consider an affordable price for such a customer. I would like to suggest something like $50-$60 USD a year (discounted for paying annually) with something up to 1gb of space.

Thank you.

@soulless: you talking about just offering one one plan and one price? What does everyone think of that?

@sol, I really like your idea, and would suggest a slight change. I'd recommend providing 3 kinds of plans: a free, limited plan; high-end premium plans with additional services at fixed monthly prices for more intense usage; and a median pay-as-you-can plan, as sol recommended, for the rest of us. But I'd add one more twist on the pay-as-you-can plan: place reasonable maximum prices as well. Once you reach a certain price point, you're essentially paying premium plan prices.

@Ariejan, I completely disagree with you and find your criticism (in your blog post) to be highly inappropriate. Respectfully, you don't know my business, you don't know my requirements, you don't know anything about my experience and you certainly know nothing of my financial situation. Worse, your assumption that I and others are asking for low prices simply because it's the new tool on the block, is presumptuous at the very least. Joel Moss asked and I answered based on my needs. You're free to provide your own assessment, as well as disagree with me, but it is not your place to belittle my assessment without knowing a lot more about me.

By the way, I justified my answer on additional grounds as well, including competition with ProjectLocker and the basic principle that distributed SCM intentionally facilitates additional forking, branching and versioning, to which you gave no reply.

@Joel Moss

I think most of the comment's above echo the same sentiments with regards to artificial limitations of number of projects and repos or feature set (they all hate it =).

If you don't artificially limit things like projects or repos or feature set etc... then you don't need to form different plans. The only thing that differs from each customer is really the amount of resources used. For example, projects or repositories or features don't cost money, space or bandwidth are the things that costs money.

So one simple low price with an option to add more space (at additional costs) works out the best for everyone. Throw in a discount for paying annually instead of monthly (less payment processing fees) and everyone saves money too.

Whereas other companies like codespaces and unfuddle artificially limit too many things for no good reason. For example, a developer with 20 projects utilizing 100mb has to pay more then a developer with 2 projects utilizing 1gb of space. Its a bit silly when the companies are using Amazon EC2 where it costs money to store more data, so they are infact sucking more profits off the guy whos using less resources.

IMO repositoryhosting does a good job on its simple pricing. Everyone gets the same plan with all the features, the only extra costs are space. So small developers won't get charged for what they don't use. The costs will also scale accordingly for larger developers who will use more space and bandwidth.

So if Codaset can do something similar to then you've got my vote.

Even if there's no free plan for private repositories, I would still choose Codaset over github for storing my private repos. As long as the pricing structure is based on storage size (not number of repos or contributors).

One private project is necessary.

The reason I'm trying Codaset and not GitHub or another service is because I want to try the service with code I'm actually writing and working on, and it's not an open-source project. (The value in letting us use the tool on our actual work, instead of "test projects" we're not likely to spend much time actually testing, is considerable.)

I'm fine with paying for something if it's good and useful. GitHub is too expensive though for a hobby developer like myself when the primary thing I'm looking for is backup and I don't want to run my own server in my home. Something around $3/month for the same level of service GitHub provides would be delicious and probably pull a lot of GitHub customers away.

Please structure pricing primarily on storage used and not arbitrarily on # of users or # of projects or # of repos.
So far the most affordable option for me which gets my files in a cloud is to run everything offline and symlink my working folder to a dropbox account. I get 2GB for free and when I change files they upload automatically.

I'm not really sure I like working this way, so I'm trying out CodaSet after using Unfuddle before (I had to leave Unfuddle because prices were unreasonable)

BTW I really need an api so I can easily import my issues from my current system (SD)

Hey guys, I would love to know your thoughts on this http://codaset.com/codaset/codaset/wiki/proposed-payment-plans

@ Joel Moss

See the comment I made 11 days ago....

"Whereas other companies like codespaces and unfuddle artificially limit too many things for no good reason. For example, a developer with 20 projects utilizing 100mb has to pay more then a developer with 2 projects utilizing 1gb of space. Its a bit silly when the companies are using Amazon EC2 where it costs money to store more data, so they are infact sucking more profits off the guy whos using less resources."

Your proposed payment plans basically enforce the same thing. Under the proposed plan, the person with 20 projects using 100mb has to pay $20. While the person with 2 projects using 1gb can sit comfortably on the $5 plan.

Shouldn't the person utilizing less resources pay less? Why is the person using less resources paying so much more?

@Alexander: That is not the case here. I am not charging for the amount of resources you use. I'm charging you for the service, and the features that Codaset provides, as they are more valuable and most likely the reason why you all use Codaset in the first place - hopefully! ;)

@ Joel Moss

I think thats preciously what everyone commenting above didn't want to see happen. Charging by project.

You noticed it yourself when you said:
"So then if every one wants to pay for space and not projects, what sort of space are people currently using?"

Alot of us have alot of small mostly inactive projects that we revisit once in a while, like utility projects along with our normal projects. We want to pay for a fair service, when we have to pay more then someone else while using less of the service then them, then its unfair for us.

We were just hoping for a payment plan that didn't follow the trap of other companies by artificially limiting this kind of 'feature'.

I'm not sure what are the real costs involved between project and space on the repository provider's side.
It seems all of them are using this kind of pricing plans.

Although, I really do like the pricing based purely on usage, that's really advantageous for me.
I also think there are more reasons on why we host our source code on codaset, github, etc. instead of hosting them on regular cloud service or web hosting provider, which also provides us with the ability to create and manage our own repo.

I think if all you are looking for is a place to host your repo's and nothing more, then obviously you would only be concerned with disk usage, as nothing else is offered. But Codaset is not all about the repo, it is all about the project, which I like to think is what separates us from Github.

And because we are all about the project, disk usage is not really of importance as much as improving the experience and tools to support those projects. So I therefore place 95% of the value on Codaset itself, and not on disk usage. I want people to use Codaset because it is really useful when developing and managing software projects.

By the way guys, this comment thread rocks! I really appreciate you all pitching in with your thoughts. Gives me a warn fuzzy feeling inside. Cheers!

@ Joel Moss

We do want the project features but we don't need the space. We use notes, the ticketing features to store todo, bugs , wishlist etc... for each small project since it helps remind us of things when we go back to them every so often.

I understand if its not feasible or profitable (your a business afterall), but it feels like you have to buy 20gb to get 20 projects only to use 200mb. Which is similar to other companies that force you to buy a corporate plan in order to get more projects when really we're a small developer.

Unfortunately, I can't really think of another payment model besides charging based on resource usage. I really do like codaset but I'd end up paying too much per month just for projects.

PS. If your just comparing number of projects, I think this payment plan might be much more expensive then other competitors once you hit 10 or more projects.

@Joel Moss, codaset is only as valuable to me as it supports my needs. I prefer codaset to all the rest. But the proposed plans simply render it completely nonviable for me.

As I see it, "storage" is a resource, "collaborators" are resources, "projects" are resources, but "project management" is a service. The model you've suggested tries to charge for codaset as a service but instead charges for just another resource. Legitimate, but unintentioned.

I prefer you charge by storage because it's a predictable model I can plan for business-wise: it's a linear expense, kind of like paying for office space. But projects are a tool of the trade, a dime a dozen, created for all kinds of different reasons: it's like charging a toolsmith extra for what should be a cheap hammer. Worse, it's an imposition: codaset is telling me how to do my job.

If you want to charge on the merits of "codaset as a service", you can charge by enabled major features per month (blog, push api, ticket system...), or make all the major features available to everybody but offer different levels (e.g., flexible premium level, non-configurable basic level, etc.), or charge by total throughput (which is a resource for you but part of the service for me), or offer "sub-leasing" similar to Campaign Monitor so my clients pay you for extra features directly. I'm sure there are many more alternatives.

In short, the proposed plans are simply not feasible for me because they don't work the way I do.

Well that is the problem every business owner faces; trying to offer plans that work the way you do. Unfortuantely that will never be possible. There will always be users who love what you do, and users who don't. What I need to do is try to be as flexible as possible.

I think the proposed "pay as you go" model is very flexible, as you pay for what you use. As opposed to paying for a set plan that lets you create up to 10 projects, but you only have 3.

I understand and appreciate your concerns Noam, but I have to make a decision that I think most users would prefer. I am also not a fan of restricting what features are available. I want everyone to be able to enjoy everything that Codaset has to offer.

I really like the "pay as you go" model; it seems to be the fairest to your customers. However I strongly dislike the idea of a minimum monthly charge that is significantly higher than the base price per project.

I'm a hobbyist developer; I work on projects alone and I have no budget. Codaset is great, but it seems a bit costly to charge a minimum of $60/year for paid features. (Hosting my website has cost me less than $20 since it opened in 2007, and that's including getting Slashdotted in mid-2008.)

Having a minimum charge tells developers like me that we're not welcome here - it's the same message that Github sends: "if you're not willing to pay a lot, this place isn't for you." I realize that to a business $60 is nothing, but as cool as Codaset's features are, $60/year is really pricey for someone who codes for fun, doesn't make income from coding, and doesn't need any of the collaboration tools (since they're an individual developer.)

It seems like the equivalent of running a restaurant and charging every customer for a meal, even if the person just wants a snack.

I have no further comments on the pricing, I do like pricing based on storage, but I could afford the pay as you go plan.

@chardish:

I think if you're really working on your projects just for fun, why do you need to keep it private?
I won't pay Github, if I just want to host my fun projects. I'll just create a public repo.

I think Github and Codaset main message is social coding.

@Jesse Armand:

For the same reason that if I were a writer who were writing for fun, I might not necessarily be comfortable making my writings public, particularly not the revision histories of my writings! Some things are simply works in progress : )

Well if you're really writing a code that you will eventually need to protect, that's not really just for fun.
I would rather just store it myself, if it's really that important.

In other case, if you're writing something that's currently in the works, imperfect, or something like that, and if you really need it to be hosted here, with all of these features, I can't see the reason why you need to protect it.

I don't care if I upload something that's useless on github, cause nobody cares, they won't just look at my repos and consider me as an ugly developer or something like that. I saw lots of inactive account or empty projects.

Noam's reason is more reasonable, but I don't think yours is ;-).

Not surprisingly, the people who are unhappy spoke up first and loudest. I just want to mention in a quiet voice that the proposed pricing will be cheaper for me than GitHub's, so I've got no objections.

@Jesse Armanad:

I have a fair number of projects that I keep private for the following reasons:

• they were old university work and I know the school recycles variants of the projects
• (old) server configurations that I haven't scrubbed for private data
• projects for which it's just easier to include private data (password, AWS key, etc.)
• projects which I haven't properly set up for opensource (lacking copyright statements in headers, no README, no LICENSE, etc).

I contribute to a lot of OSS projects. For me to just open up every repo I have would dilute the value of my open repos because it would be harder to sift through the ones that are legit and those that are not.

Having said that, the pay-as-you-go model looks like it makes the most sense. It won't work for me, so I'll just stick to hosting my personal stuff privately. But, the pricing model works great for my small business, so I'll roll with it.

@Jesse Armand:

You raise an interesting philosophical discussion, and you raise some valid points, but programmers come in all different kinds of varieties, people have different attitudes about their work, and not everyone is willing to declare "yes, world, you can have the source to this forever" before a project is done. (And you can't discount the possibility of bots archiving public github repos for purposes of preservation - once something is public, it's public forever.)

Just because I don't think my stuff is worthy of protection doesn't mean that I'm willing to let the whole world see it. I see it as more of a privacy issue than an issue of paranoia or greed.

Well, okay no problem with me.
I hope Joel Moss could figure out what's best.

Dosn't seem to be a way to get off the comment updates? There dosnt seem to be an unsubscribe option.

(Reposting: the comment above lost formating so trying again)

I am right now paying $5 pm. for RepositoryHosting.com.
It gives me --

-- Git

-- Unlimited number of projects

-- 2 GB limit.

-- Trac (I hate it)

-- Poor synchronization between site-login system and the Trac login system.

-- SSL

-- Fixed width website layout, not changeable (I hate fixed width layouts)

-- Project names support dashes (-), underscores (_) and dots (.).

You might think the last line to be funny. But it is shocking how many project hosting sites out there won't support dashes and dots in the project names (ProjectLocker doesn't; I believe Unfuddle doesn't, so I don't use them).

I am eagerly looking for some other host which could give me the following:

-- Git

-- Special characters allowed in project names

-- Redmine (this would be an AWESOME PM system worth paying for)

-- 2GB space

-- unlimited repositories

-- Under US $ 3 per month.

-- SSL.

-- Multiple private users with different access rights to different branches of same project.

-- Project statistics (number of downloads, number of forks , project activity, developer churn rate, bug/feature tracking on integrated Redmine, etc.

-- Customizable site layout (with many fluid width layouts)

-- Site APIs (a MUST these days; for Web2.0 style project tracking)

-- Associate a hierarchy of project tags (different from Git tags, e.g. programmming languages use, target audience, Industry segment etc. as on Sourceforge)

Provide something like this and I would switch in a heartbeat.

Wow! You're not asking for much are you ;) I am sorry, but $3 a month for unlimited usage of Codaset is simply not feasible, as I am trying to run a business. If I were a non-profit, then maybe I could do that.

Let us look at this "modularly":

1) Space and bandwidth.
Webhosting companies go as low as charging $1/mnth for 200GB/unlimited bandwidth.
It shouldn't matter how many projects/users etc. you put on that.
2) Services.
This is basically all the rest like trac, SSL, git, etc. etc. (in other words, every single feature.)
You could charge a minimum of $1 per feature used as well.
3) Pure profit.
Simply charge $1 per account for yourself.

So, this way people can create their own package of features starting prices at $3 for a basic starters package consisting of:

200GB
unlimited projects
unlimited users
git

Then people can simply add feutures as they go each costing (a minimum of) $1. You could limit the maximum price, or not.

How does that sound?

You should take a look at this blog post and its comments http://codaset.com/codaset/codaset/blog/which-business-model-do-we-go-with.

All I ask is that you remember that Codaset is not a web hosting company, and disk space is not the only thing that has value. I've mentioned this before, but Codaset is all about the feature set, and not the disk usage or code hosting. Codaset itself is worth much more than disk space ever will be.

I know, that is exactly what I am trying to say here.
Simply charge for each feature. The total cost of the features will always exceed the price for disk usage/bandwidth. But, at the same time you can also cater for those of us who just want the minimum set.

If the argument against this would be that many would go for the "cheap" option (e.i. just disk-space and git) than you should reconsider your philosophy. But personally I think that if you offer good features most users will grow into your product and start paying more as their needs grow. While at the same time you don't shy away people who cannot afford the more feature-rich packages.

From a marketing perspective this would give you a definitive edge as your users can create their own package depending on their needs and budget as opposed to other services where you always need to choose between fixed packages with fixed prices. The downside to this usually is that the basic package is too expensive and that you pay for features you may not use/need.

Something else I just realized, if you could set this up like this, you could also introduce pre-payments. Have users pay an amount in advance which they can use up depending on the features they use per month-basis. But this is just an extra thought, probably too complex to implement at this stage.

All good suggestions, but a little too late I think, as I have pretty much decided on a pay as you go model as described in http://codaset.com/codaset/codaset/blog/which-business-model-do-we-go-with. Thx anyway.