Auto-Canceling Subscriptions
blog.beeminder.com
blog.beeminder.com
We do try to be clear about this policy up front:
> If you stop using Beeminder, we automatically stop charging you. We only want your money if you're actively getting value from Beeminder!
And here's a thought experiment: Suppose you subscribe to the print version of The Atlantic Monthly -- like actually delivered to your house (stay with me) -- and they have a little tracker that tells them when you crack it open. If you didn't ever open last month's issue when it's time to send the new one, that's it, they don't send it and don't charge you again.
Even with the magic tracker, that's probably no good because you might want to collect them. Not likely but they probably shouldn't presume you don't want what you explicitly paid for since it has possible value whether or not you open it. I know one startup (http://serps.com) with a similarly watertight excuse not to auto-cancel subscriptions: SERPs constantly fetches and compiles data for you. You can take a multi-month hiatus and still be accumulating value that you'll use whenever you come back. (It also costs SERPs money each month to do that.)
But for the rest of us there's really no way the user could prefer to keep getting charged without any possible value accruing.
1. You are shooting yourself in the foot, revenue-wise.
2. The cognitive overhead of figuring out this unexpected pricing scheme is going to scare away potential customers.
It smacks of white knight syndrome to me. I understand that your service is trying to help people, but I don't see how this pricing scheme adds additional value. If (and this is a big if) you managed to spin it into some kind of publicity, then it could help. Otherwise I'd guess it's not going to turn out at all well.
1. It may be money you're leaving on the table but you can't really say it's money you're earning. It's like a laziness tax on your users. (And, uh, not in a good way.)
2. Your point about cognitive overhead I think is quite backwards. The beauty of auto-canceling is that it's less cognitive overhead, other than the fact that it's unusual. (I hate it when "it's industry standard" is the excuse for not doing the right thing.)
There was another Hacker News thread about pay-as-you-go vs tiered pricing: https://news.ycombinator.com/item?id=5664512 . I think the cognitive overhead of perennially re-evaluating whether you're getting your money's worth is a big reason developers like pay-as-you-go. And the biggest reason you wouldn't be getting your money's worth is that you're not using the service at all anymore.
So I think auto-canceling subscriptions are the best of both worlds: simple and predictable, you can budget for it and think about it in terms of the value you're getting (an ongoing service) without the risk of paying for something you're not using. How's that not clear additional value?
I disagree about a standard "you need to cancel it" model of canceling being a laziness tax, but I can certainly see reasonable people disagreeing on that point.
I should have explained my thinking on cognitive overhead a little more clearly - I think the overhead of explaining it at the beginning of the subscription is what could turn people away. No disagreement from me that once a user is signed up, they don't need to think about it as much.
I certainly wish you the very best with this billing experiment, and will be excited to read the results if or when you choose to publish them!
charge(user, amt) unless last_active(user) < time() - 1.month
Imagine a friend/friend version of that business/consumer relationship. You would not keep taking your friend's money!
Anyway, yeah, I think we're on the hook now to follow up on this in a few months or whenever we have enough data. :)
Most businesses don't approach subscriptions this way. Some portion of revenue comes from people choosing a "set it and forget it" approach toward service subscriptions. Is it unearned revenue? The traditional take is the user has access to the premium product. If they don't use it, that's their problem. (i.e. a gym membership)
I think traditional MBA thinking is to optimize around revenue at any cost. However, I think this sort of behavior has made people, (especially young people), not trust businesses from banks to wireless carriers. I think businesses that optimize around revenue rather a relationship with their users are likely to lose them as soon as there is an alternative.
Clearly, it is in the best interests of the user to do this though, as a business you should also expect to drive more long term loyalty to the product.
I think the entry could make an argument that this will cause users to be more loyal to you or create longer lasting subscriptions. Without data or even anecdotal feedback that users are more likely to stay because of this choice, it sounds like you are something of a White Knight turning away money.
From a user perspective, I'm used to a use it or lose it subscription. So, if I am not charged, and then I all of a sudden am charged because I came back, I might get confused. Do you do customer education at the time of the subscription creation of this friendly charging policy? Do you notify the user by email when they are not being charged because of a lack of activity? (This might be a good way to actually get them back to the product, and possibly you have measurable data to support the value of this?)
If there's no question that auto-canceling subscriptions are in the best interest of the user then I feel like it's case closed.
As for customer education about the policy (repeating myself from my other comment) we do have it in large print -- "If you stop using Beeminder, we automatically stop charging you. We only want your money if you're actively getting value from Beeminder!" -- but you're definitely right that we should email the user when the auto-canceling happens. I think that could generate a ton of goodwill and bring them back, as you say.
As far as auto-canceling subscriptions, the gym could also make an argument that they have to provision a lot of physical space and equipment based on how many people have paid to use the space. They don't recover that cost when you don't happen to show up.
None of this applies to most SaaS companies though!
But, yes, any startup using pay-as-you-go is exempt from my objurgations. :)
If automatic data entry is enough to be considered active, maybe you should warn people that the data points entered automatically will make them count as active.
The more I think about it though, if you're not signing in at all then you'll almost surely derail before long, which means your graph will freeze and you won't be active any more, automatic data source or not.
In fact, it now occurs to me that this is an example of when it is fair to charge you for a month even if you weren't actively paying attention. It's like my example with SERPs.com in another reply here. If Beeminder is actively collecting data for you and storing it and graphing it then that's value that it's fair to pay for.
If you do entirely wander off then even with an automatic data source -- unless you set preposterously unambitious goals (don't do that!) -- you'll end up marked as inactive before long. So I think the spirit of auto-canceling is still being honored.