I got flamed by users that felt we didn’t give them enough advance notice about changing the subscription plans and restricting free private spaces.  The comment thread from my last blog post was long and educational.

In penance, I will try to clarify a few things here.  I will also summarize the discussion, for people that are interested in how Internet products get priced.

First, this is not an emergency.  We will provide a transition period of at least a few weeks for free-private users to upgrade or migrate.  If we do not hear from you in that time, we will not expose your data, or delete it.  We might restrict it to read-only at some point.

Second, the minimum disk charge is currently $0.30 for 100 megabytes.  So, the minimum charge for a space is $2.30.  Actually, if you read on, you will learn that we are capping the user charges, and the minimum space charge has been lowered to as little as thirty cents!

Third, a "payment unit" pays for $1 of subscription charges at full price.  Some of you have opted for advance payments of $100 to get 120 payment units, and I thank you for that vote of confidence.

Fourth, if you are an existing subscriber, you don’t need to do anything, and you will get a price reduction in most cases.  We will credit the full amount of your last payment as "payment units" for the new plan.  If you think this will cost you money, we can provide additional credits, or a refund.

A few people asked if their team members and customers will have to pay something.  Fortunately, in our system, the subscriber / space owner pays for all of the team members.  The team members that you invite don't pay, and they never have to submit payment information.  We never ask team members for payment information, unless they go to create their own space.

Price Reduction Again!  Per user charges limited to $8

Our users are negotiating in various ways for lower prices.  I was not sympathetic to many of their arguments because $2 per user is a VERY low price.  It needs to be higher to support reliability and features at the level of quality we all want.  The negotiation never seems to end.  Two years ago, we were charging $50 per space per month, and users said that was too much.  We reduced it to $19 per space per month, with a discount to $12.50/month for annual purchase.  People bought the annual plans, but others said that the $19 number was too big, and they wanted a way to buy smaller amounts.  So, we broke it down into users and disk space, and now we sell a space for as little as $2.30/month.  That sounds like an amazing deal to me.  There is a limit to how low we can go and not be wasting our time.  I have four kids, and employees that rely on me, and they need to eat.

So, it was with a hard heart that I first looked at the analysis from Joao Saleiro, who has six users on his team, and brings them all into 10 projects.  That would add up to $120 under the per/user per/space plan.  Other people with a lot of projects also noted that their charges would escalate quickly.  Joao is exactly the kind of power user that I want on our system, and he’s using the space structure the way I think it should be used.  I looked at my own list of more than 60 spaces.  He was right!  We were charging too much for users with lots of spaces, users like me. I crumbled.

We are making an important modification. The price per unique user will be capped at $8 per month per payer. After you add yourself or a team member to four spaces, the fifth space through infinite spaces are free for that user. If you have the same team in many spaces, the extra spaces cost you only storage charges (30 cents, in most cases), and premium tool charges, if there ever are any.

This means pricing of $8 per user for serious users, and you can forget the "per space.” You won’t have to worry about archiving spaces, because the extra spaces won’t cost you anything.  And if you are frugal and have only one space, you can get going for $2.

And, if you pay in advance, you get EVEN MORE DISCOUNTS.  If I sound crazy, it’s because I can’t believe it myself.  We’re open for business.

Upcoming communications

Other suggested pricing schemes

Many of the commenters, persistent negotiators, proposed schemes that would reduce the prices that they would pay.  We have adopted the most important suggestion – a limit on per-user charges.  I will review the rest here for your comments.

You should show advertisements
Good idea.  I’d like some advice from an expert about how to make advertising work for us, if any of you want to provide it.  However, it won’t solve our problem, because the fast usage growth is with subversion clients, which won’t render ads.  And, we are talking about about private spaces, which have limited visibility anyway.

You should provide free services to us students.  We don’t have a good way to pay.
I want help students.  I like the student projects on Assembla.  We tried providing a special free student plan, but a LOT of people emailed us to request it, so it became costly to administer.  I had to work on it nights and weekends.  We will try to arrange something for them that will be more automated.

You should sell us a fixed price bundle of users, spaces, and disk quota so we can get more spaces or more users for the money.
This is the Basecamp/Unfuddle pricing strategy.  It would result in a price reduction for people who have the right mix of users and spaces.  We considered this, but we rejected it for a very specific reason.  We want to keep expanding our feature set and adding new “tools” like instant staging servers, which will be amazing but will cost extra money.  With fixed price bundles, you can’t add new services without breaking the fixed-price model.  What actually happens with vendors that use this model is that they put new features into new products.  I have talked to a lot of users of these fixed price bundles, and many of them are spending more than they would spend at Assembla, because they buy more than one product.  This is not only expensive; it creates user management hassles.  So, we went with “metered billing” – you pay for what you use at the end of the month, rather than a fixed price at the beginning of the month.  We give up a lot of cash flow to do this, but it gives us the flexibility to include new services. 

I think our $8 per user cap gives you the best of both worlds.  You can start as small as $2.30.  If you more fully adopt the product, the price is essentially fixed at $8 per user, so you would pay $48 for a six user team, and you can use an unlimited number of spaces.  There is no need for “archiving” or any other bothersome space management.

The user charges are too high.  You should give us a lower price for “client” users.
We are already at $2 per user.  How low can we go before it gets ridiculous?  My competitors are charging $50/users.  And, haven’t we seen that users will always put themselves into the cheapest category?

You should charge only for disk space.  I don’t use much disk space.
We considered this, but we discovered that most projects use less than 100MB.  If we charge only for disk space, it would need to be a high charge, which is not what the people making this suggestion are looking for.

You should charge only for “activity”.  That way, heavy users will pay more, and I will pay less, because I don’t use my spaces very much.
If we try this, we will have a new argument about what constitutes “activity”. What will you say if I charge you an extra $2 because your RSS reader keeps checking the event list?  Furthermore, this sounds like a recipe for receiving a very small amount of money.  How will I cover the $40K/month in expenses?

Was this an evil plan?

Thank you to everyone who supports us with words or a subscription.  A few commenters called us bad / evil / asshole / disappointing-bait-and-switchers for drawing in users, and then suddenly proposing a change.

Our communication has been limited, even stilted, but our intentions have not been hidden in any deliberate or sneaky way.  During the last three months, we have discussed all of the details of the new subscription implementation on our ticket list, which is visible to the public at https://www.assembla.com/spaces/breakout/tickets .  Going forward, we will continue to be completely transparent.

The restriction of free plans wasn’t some sort of evil plan.  We thought we were doing a good thing by offering the free plans with no restriction for more than two years, and we did not think about restricting the free private plans until recently.  The free services have given us a chance to work with a lot of great agile developers, which helps us with recruiting.  That is why we still offer a generous free public plan.  However, like all bootstrapped startups, we need to be adaptable.  We discovered three things that eventually forced us to make a change.

• Staffing services will not cover the cost of our software services (the original plan).  The standard gross margin of 10% just isn’t big enough to provide a profit, unless we spend a lot of VC money to get to a much bigger size.  Assembla is not VC funded.

• Many users came only for free private subversion repositories, which are especially difficult to "monetize".  I was happy to pay for these repositories as long as the basic hosting and admin and support cost was affordable.  Our total costs are around $40K month (making us a small but serious business), and our variable hosting and admin costs are around $10K/month.  However, growth of the free services is exponential at about 8x per year, which means that at a certain point costs go beyond what we can afford.  A $5K/month expense today (I’m paying at least that much personally for your free services) is a $40K/month expense a year from now (definitely not affordable). Once the exponential turns the corner, it blows us out of the water.  And, as the number of users grew, free users started calling me at all hours with support questions, which is another type of expense.

• We have a lot of commercial users that will stick with the free service as long as it is private and relatively generous. We need to bring them into the boat as customers.

So, we made a decision to compete for subscription revenue. That's great news for you, because it means we need to do an even better job on our features and customer service.