Here’s some philosophizing and strategery from a recent daily beemail. I’m going to bury the guts of it because – based on dozens of replies to the beemail – we’re pretty much convinced to drop the idea. Not totally convinced, but pretty convinced. The whole thought experiment was valuable though and I wanted to have it in the forum for posterity.
You can go ahead and answer the poll without reading the background.
Pricing experiments and sting dilution
Recently we tried a kind of opportunistic experiment with not having a free plan while hiring our newest support workerbee, Robin. For a few weeks, everyone had to go straight to Infinibee at $8/mo to even sign up. That seemed to be a bad idea. What about the other extreme? Maybe almost nothing should be paywalled. More goals = more money = more user-awesomeness. Anything (like premium plans) that throttles goal creation ultimately hurts us. It’s a theory we’re probing.
So here’s some total Sky Pie but let’s run with the thought experiment…
Kill the free plan AND kill the premium plans. Namely, after a generous initial trial period, you have to pay Beeminder at least $X/mo.
You could think of it as “you didn’t derail enough so we’re charging you to get you to the minimum” or you could think of it as “Beeminder costs $X/mo but derailments are applied as credit”. (Probably the latter is the better spin!)
It sounds like anathema but what if it’s great? We’ve always thought of “don’t dilute the sting” as inviolable but if we reject that, maybe everything else gets way better. Pretend we’re Einstein rejecting one principle (speed is always relative) that defies all common sense to reject but that makes everything else fall into place!
Arguments that sting dilution is totes ok:
- Derailing’s not failing and acclimatizing users to derailing is awesomeness-increasing.
- Derailing is kind of the opposite of failing. You derail by having ambitious goals and that makes you get more accomplished. So even literally incentivizing derails is .
- The initial $0 and $5 derailments aren’t motivating anyway. You need to get up to your motivation point. Having a derailment quota barely changes that.
- Derailing is already partially incentivized because some users (the bestest most amazing ones) want to make sure they’re supporting Beeminder. A lower bound that’s charged no matter what elegantly solves that for them.
(Does it solve it too much, from Beeminder’s perspective, if some users treat themselves to derailments generously, motivated by wanting to support Beeminder and introducing this money-grubbing $X/mo minimum makes them stop doing that?)
Having a minimum amount that everyone pays also elegantly solves the problem of – what’s the diplomatic word? – freeloaders. There’s a particular kind of user – and I think we accidentally encourage this in our marketing copy – who treat Beeminder as a game where the goal is to never ever pay Beeminder a penny. Some of those users have conversations with us in support about each derailment and why it wasn’t legit. It’s not actually common but it’s frustrating when it happens.
If you’re still recoiling at the idea of diluting Beeminder’s sting, we could have an option to pay the $X/mo unconditionally. No, wait, that violates the Anti-Settings Principle. But you could make a dummy goal that derails every month like clockwork to hit the $X/mo minimum. Or we could point you to a third-party tool that hits the “charge me” API endpoint once a month. Or, y’know, we could not be dogmatic about anti-settings and go ahead and give the option to pay us unconditionally if there were real demand for that.
Maybe it all kind of works?
Less than 24 hours ago [from floating this in a beemail last month] I thought this was a ridiculous idea that I asked Bee to help me think through just as an exercise in case it spawned any real ideas. But now it’s actually growing on me, and Bee’s not hating it either. User-me is big on blog.beeminder.com/defail and I’d be happy to have pledges applied to premium and not mind at all a feeling of the first N derailments each month being effectively free.
Does that mean it violates the Anti-Leniency Principle? Well, less egregiously than the examples in the anti-leniency blog post. At least we’re not tampering with the concept of derailing, to make it more lenient. There’s still a single crisp bright line. Rather, it violates anti-leniency similarly to how starting with $0 at risk violates it. This does suggest that if we did this, we’d want to be very clear on where you stood on hitting that $X/mo minimum. The principle is to minimize fuzziness and ambiguity about the consequences of derailing.
Ok, that’s where that idea’s at. It’s drawing heavily from user-me’s feelings and I’m a weirdo, so I’d love to hear your thoughts!
- Sounds nice!
- I would opt out of that. Don’t wanna dilute the sting or feel partially incentivized to derail!
- I already feel partially incentivized to derail because I want to support Beeminder.
Approval voting, as always. Vote for all that apply.
The results from the beemail replies and the Beeminder Community Discord (combined) are like so:
- 23 "sounds nice"s
- 14 "opt out"s (or "don’t even give me that option"s)
- 9 "already incentivized"s
You’re encouraged to vote again in this forum poll even if you voted via beemail or the Discord.
And here are very paraphrased and very incomplete excerpts from the dozens of beemail replies but I encourage you all to repeat your full replies as comments!
Beemail reply excerpts
- “I think it’s a terrible idea [+ comprehensive tearing to shreds of all the arguments in favor of sting dilution]”
- “The one-time yearly premium payment is (usually) far enough away that it wouldn’t really affect my feelings about derailing (as long as actual money comes out of my actual account when I derail)”
- “Sort of does dilute the sting for me”
- “I don’t want extra mental management overhead”
- “Incentivizing derails across the board isn’t actually awesomeness-increasing”
- “I don’t find most of beeminder’s value in derailments; I find a ton in the way rerailing is handled, compared to almost anything else I’ve seen out there. The combination of automaticity, buffer, etc is pretty great. I’ve never been particularly into the sting aspect, though it’s possibly growing on me slightly over time.”
- “For me ‘Beeminder is a game where the goal is to never ever pay Beeminder a penny’ is kind of the point”
- “This sounds like the exact opposite of what I want; I would be happy to pay an initial $ amount to create a new goal”
- “Oh no, now I’m worrying I’m one of those users who has conversations with you in support about why every derailment is not legit and never pays for them [ironically said by the exact opposite of that kind of user]”
- “No, no, no! Agh, no! No, no, no! This would destroy Beeminder, or at least Beeminder-as-I-now-use-it-and-value-it”
- “My gut reaction as a student was yes!”
- “I’ve always kind-of felt this was the right approach. For me, derailing is the sting.”
- “I don’t think it’d figure into my reactions to an imminently derailing goal”
- “If derails no longer cost money in some situations, then, in my mind, Beeminder is not doing its job”
- “Having the service cost a certain amount each month or year makes sense”
- “If you do that, please please implement a way to cry uncle on a goal first”
- [A couple Beemium people emphasized they don’t want stings (monetary commitment devices) at all ever but Beeminder needs to not be focused on that use case – Beeminder is allll about stings]
- “Noooooo. As soon as you have a set number of ‘free’ derailments per month, now that’s a resource that has to be used or lost.”
- “I buy the argument that this would help get people used to routine goal derails as a good thing. Even after five years, I’m pretty derail averse even though I totally buy the derailing isn’t failing argument.”
- “The framing changes how I feel about it entirely even though it’s functionally equivalent”
- “What if you give people imaginary currency for each dollar in their monthly budget that they paid to you but not out of derailment fees?”
Again, my conclusion so far is that this was a fascinating thought experiment and generated some profound, insightful, and passionate feedback and the conclusion is that we don’t want to actually do this. I really appreciate being able to think it through with you all!