Our offer was so good it backfired.
Last month, we officially launched BigStash. In short, BigStash lets you store your infrequently accessed files in a cheaper (but slower to retrieve) service, ideal for archiving purposes. It’s the kind of service usually offered to the enterprise, but we believe there is a significant number of consumers who need something like this too.
We decided to go ahead with a big offer, something that would make us stand out in such a competitive market.
As expected, it worked and we got a lot of traction just because of it.
What we did not expect, were the negative comments! “It’s a scam”, “2 weeks later, company and domain name doesn’t exist”, and “Sponsored by the NSA” were some of the comments we got on our Facebook page. Because, to a good number of people this sounded too good to be true: They thought there must be something we are hiding.
So let me tell you the inside story, and you can decide if offering 5TB to every user, for free, for one year, was a good business decision.
1. How much will it actually cost us to offer everyone 5TB for free, for one year?
We are using Amazon’s AWS Glacier to store user data. That’s $0.01/GB/month (plus a number of other complicated charges, per request, retrieval bandwidth, etc, but let’s leave them aside, minimizing these is what we’ve been working on for the last year). In simple terms, this means: If you upload nothing, your 5TB will cost us nothing. If you upload 5TB it will cost us $600/year.
So, the more each individual user is using their year-long trial offer, the more they cost us. On the other hand, we have a good reason to believe that heavy users (the ones that cost us more), are more inclined to stay with us after the trial, and that, given the nature of our service (long term storage), they will stay with us long enough to amortise the initial acquisition cost.
And yes, the $500/year we charge for a 5TB plan is less than what it would cost us to store 5TB for a year, but if this is a popular usage pattern, we should probably offer even larger plans.
Which brings us to the next point.
2. We need real usage patterns.
Consumer cloud services pricing is largely influenced by usage patterns. Low-level PaaS services like AWS expose every single cost involved. They are great for developers and the enterprise but not for the consumer market.
Take for example AWS Glacier (BigStash makes heavy use of it). To calculate actual costs, one will have to take into account storage costs, the cost per request (including a hard-to-decode note that: “You can retrieve up to 5% of your average monthly storage (pro-rated daily) for free each month. If you choose to retrieve more than this amount of data in a month, you are charged a retrieval fee starting at $0.01 per gigabyte. […] In addition, there is a pro-rated charge of $0.03 per gigabyte for items deleted prior to 90 days.”) and data transfer costs.
But most consumers don’t like doing complicated calculations to find out what they will actually pay. They want a simple pricing model, associated to something they understand and they can control (like the space they need). It’s up to us, the service providers, to estimate and aggregate all hidden costs, and offer a simple and compelling package.
So, we need these data points: We need to have a big sample of actual usage patterns that will help us decide if our pricing makes sense. I would take it even further and say that if we could pay to get usage pattern data, we would do it —our business depends on it.
And, in a way, that’s what we do: We are willing to invest a good part of our funding to offer the service for free to a large number of users, just to make sure we have good data to prove our business model.
Two weeks ago, I sent an update to our investors: I was able to dig into data, cluster usage, identify usage patterns, do cost and profit estimates based on tens of thousands data points. Just for this, it was worth every GB of storage we offered for free.
3. We need Beta testers.
BigStash is still under heavy development, and every now and then we will bring the site down for a few minutes, or make a change that will break something until we fix it again and so on. That’s what being in Beta means, and we don’t hide it.
In other words: The service is not yet up to the quality standards we would like it to be, and it wouldn’t feel right to charge users before it gets there.
The way we see it, our current offer is just returning the favour our current users do us —when they test on various local setups, environments and use cases, and when they help us find bugs and UI/UX shortcomings.
So? Do you still think we are crazy to offer so much space for free?
To me it sounds like a good business decision, but I’d like to hear your thoughts too. Leave a note here or tweet @vrypan.