I think that the idea of people having a single motivation point is a
mistake. My energy and motivation levels are in general incredibly
variable, and my responsiveness to beeminder style commitment devices
varies accordingly.
I’m definitely in general not the sort of person who can get shit done with
Willpower on average, but I go through months at a time where I’m within
spitting distance of that, and just not letting a goal derail is motivation
enough.
Right now… well my goals are currently coming tumbling down around my
ears. I’ve derailed on a few things because I valued the pledge less than
not doing the work to derail. I think those were mostly $5, but one might
have been a $10. I put my pledge on coffee up to $90 because that’s silly
money for a coffee which I would never pay and right now I’m considering
whether it’s worth it (I’m hanging on to “not worth it” by the skin of my
teeth).
Also, FWIW, I was one of the people who claimed to not be motivated well by
money. I still maintain that, it just it turns out that what I really meant
is that I’m not motivated well by monetary rewards and I can still hate
spending money on trivial things (I had not previously realised this about
myself).
My recent “Type Bee Personality” post [1] described sort of the
platonic ideal of a user of a commitment device tool like Beeminder or
StickK or whatnot [2]. In the case of Beeminder it’s someone who
climbs the pledge schedule till they find their Motivation Point [3]
and then Beeminder simply works motivational magic from then on. When
it’s too expensive to derail you make sure not to and all is roses and
ponies.
The other extreme – to take it to an extreme extreme for
gedankenexperiment purposes – might be someone who’s clinically
depressed and thinks “to hell with everything, I’d literally rather
kill myself than deal with this seemingly simple task right now.”
People who just get shit done with Willpower or something are an alien
species as far as we’re concerned. They’re not part of this
motivational spectrum.
The question: How are people distributed on this spectrum of
motivatability? Sometimes people claim to not be motivated well by
money and I always suspect that what they mean is that they haven’t
found their Motivation Point. But they could also be kind of
incorrigibly irrational and/or have a somewhat self-destructive
utility function where a commitment device will do nothing but inflict
more damage on top of the akrasia.
I want to get a better handle on that question. Could y’all chime in
with personal anecdotes? (Publicly or just to me and I’ll compile
responses.)
Well so I think I am also one of the people who’s poorly motivated by money
and much more motivated by the idea of avoiding “failure”. I think that’s
why, in general, my goals have stayed pretty low over this past year. I
agree also with the other response in that my motivation is rather
variable. I know that I definitely have had days where nothing can motivate
me because I’m in a state of nihilistic self-destruction rather similar to
the extreme example you describe.
I know, though, that I’m probably an outlier in that regard. I’d say my
average day, though, is that I have a lot of trouble believing that
anything I do is actually worth it and that I’m taking advantage of my own
rather irrational fear of failure to curb my irrational feelings of
depression. I don’t, however, hit emergency days very often since I trimmed
my total set of goals down to a manageable level. My usage pattern tends to
be that I trot along motivated for awhile and stay ahead of goals until a
bad phase hits me again and then I stop trying anything for a few days, I
start to see more blue and orange on my goals list, and then I finally make
myself start working again because I feel the pressure and after a few days
of riding the minimum of what I need to get done I get momentum back and
get ahead of everything again. I have a feeling it’s probably obvious in my
data when this happens.
I agree that a lot of what motivates me is psychic pressure that derailing is bad. I find the dollar amounts most motivating on orange / blue days when I think “Jeez, I don’t want to leave myself in a $30 eep day tomorrow and it wouldn’t kill me to [do thing].”.
If I ever get up to $90 or $210 on a goal I would have more data about whether the big amounts scare me off. So far, the two goals I’ve hit $30 on are actually goals for things that already cost me money, so it is easier to rationalize. As in, “Would I still want this beer if it cost $16 instead of $6?” is a way to sweep the $10 level under the rug.
Some of you saw that in the daily beemail I turned this into a numerical question, quantifying the spectrum like so:
0 = clinically depressed and fundamentally unmotivatable because you don’t even necessarily want to live so you sure as hell can’t possibly care about getting charged $810 or whatever.
10 = the perfect platonic ideal type bee personality.
5 might be someone who Beeminder works fine for for some things but for other things you dig in your heels and weasel or irrationally pay stupid amounts of money.
I’m a 9 I guess. I’m a lazy ass but if get up to enough money at risk I toe the line and do what I have to do, which is how I’ve accomplished pretty much everything I’ve accomplished, starting with getting my PhD thesis written when Bethany invented the “Voluntary Harassment Program” that eventually evolved into Beeminder.
I think I might be close to the “platonic beemindee” almost all of the time. I stay mostly in the steady akratic state, finishing only those goals that are eep days. Once costs get high enough, I stop derailing ($270 is the upper bound, but recently even $30 has been enough.) at all; before that, I derail every few weeks or so.
What keeps me at an 8 and not a 9 is the fact that I have days on which I am very tired (usually a lack of sleep from procrastinating on beemergencies for a week or so), so I start to weasel a little… I’m not proud of this, but recently, I changed the time zone so I would have a few extra hours to finish something (I wanted to sleep first). My hatred of messing up the QS aspects and the sanctity of my data prevent me from any sort of outright falsification.
(Speaking of which, I need to request that my time zone controls become locked.)
This depends a lot, for me, on how well I can manage the things that need doing. If I miss that something was due and only see it late in the day, I often end up closer to the “3” end of the spectrum come late evening. If, on the other hand, I see everything early enough in the day and can keep it in sight at the relevant times, I’m closer to an 8 or 9.
I am one of the clinically depressed people that you mentioned. Its very
hard to find motivation to do much of anything some days, and I’ve derailed
a lot and just given up/archived goals when I felt like “it wasn’t worth
it” anymore. I know thats not the best way to go about it, but there
definitely IS something about putting in a little data point each day, even
if I feel really crappy, that gives me a sense of accomplishment for the
day. I try to set my goals low, and then by using the “just get started”
idea, it helps to where I’ve done SOMETHING. Some days I don’t do too well.
Some days I barely want to get out of bed. But I am motivated by data and
quantification, so that helps sometimes. I’m currently on a beeminder
“break” (I know…I’m hardly the “perfect” beeminder candidate but I guess
what I’m trying to say is that I have found more use with it than I would
have without it, and if that means I use beeminder in an on-again-off-again
sort of fashion, then thats what works for me.
To place myself on the scale of 0-10, I suppose you could say I’m a 1 or 2.
Probably closer to a 1. It’s useful to give me something to strive for
that day, but if I get too discouraged by a goal I usually derail once or
twice then archive it
As someone who has suffered depression on and off, I will say that
Beeminder is not very useful, and maybe sometimes harmful, for a person in
that state. If I’m being honest it’s probably one of the main reasons I
stopped using Beeminder.
I’ve never been at “I would rather kill myself than do this thing” – in my
experience this is not how the cognition goes[1] – but there have been
times where my brain will go to great lengths to avoid even thinking about
a thing because of the hamster wheel of self-loathing that it will put me
on. In that state Beeminder tasks can be a part of that cycle of
destructive self-talk, something along the lines of “if I can’t even do
simple task x, how will I ever get my shit together?” It also doesn’t stop
me from doing the task poorly (or what I irrationally believe to be
poorly), or fudging the data.
Beeminder can be good in the short-term to get from “agonize about the
thing” to “just do the thing already” but after a while it loses its effect
for some reason. Staying on the road seems less important after it becomes
routine.
And money has never been a very effective motivator for me either
Isaac
[1] I will stop short of saying that this is a fundamental misunderstanding
of how depression works because maybe there’s someone in the universe who
has had that thought and meant it? But the numerical scale doesn’t make
sense to me either. It’s like asking “on a scale of triangle to bumpy, what
color would you say you are?”
Thanks to everyone for sharing your thoughts on this thread. I really do
appreciate hearing others’ views on these topics.
I, however, don’t really understand the scale. It seems to measure
someone’s likelihood to respond to their current pledge in the final point
in time they have to meet their goal before it derails. However, it seems
like this is actually a binary classification. You meet your goals because
you care for the pledge you are about to lose. You don’t meet your goals
because you don’t care for the pledge you are about to lose (e.g. the
pledge isn’t high enough, you have competing commitments, or your inability
to do anything via depression). Perhaps the scale is measuring how beyond
caring someone is mixed in with with how close the combination of all of
these goals are to each of their motivation points??
E.G.
Not depressed
Goal 1: Pledge $5 / Motivation Point $30
Goal 2: Pledge $90 / Motivation Point $90
Goal 3: Pledge $5 / Motivation Point $30
= Ranking on scale: 8
E.G.
Overwhelming inability to start
Goal 1: Pledge $5 / Motivation Point $30
Goal 2: Pledge $90 / Motivation Point $90
Goal 3: Pledge $5 / Motivation Point $30
= Ranking on scale: 3
My own anecdote is that I’ve just failed on two goals. (When my bank gets
its $*#@ together and finally send me my replacement credit card, I’ll
start them up again.) I failed because I sat on beemergencies days for at
least one of them for about a week and then didn’t manage to get the work
done one night as I didn’t finish normal work until 11pm. When one failed,
I let the other fail because I’d had enough and needed a rest. I was
leaving things far too late (completing the minimum required with minutes
to spare) rather than spending time earlier in the day (i.e. planning!).
One Beeminder blog post talks about setting up a goal to motivate getting
beeminder goals ticked off earlier in the day. Perhaps I should be looking
into that. Perhaps it’s just that I need to fail a few times until the
credit card debt (Huh, I only just realised that was the actual motivation
of each pledge) starts to kick my ass.
I don’t know where I currently am on your scale Daniel.
Money is indeed a poor motivator, because no matter what it’s a future
reward (or more specifically, a set of future reward options). Juice is a
reward. Facebook is a reward. A skydive is a reward. Money is simply the
possibility to get a reward, again and most importantly,
sometime-that-is-not-now. The closer the motivator in temporal proximity to
the desire action, the better.
I’m actually not at all sure the scale makes sense. The idea for that straw poll was related to our “10 freebees for $1” idea and how that’s different from unlimited freebees with Plan Bee.
The difference, repeating myself from a recent daily beemail, is that a freebee is a goal that starts at $0 but still climbs the pledge schedule each time you derail. You can turn off auto-increase and have it stay as low as $5 but you can’t have it stay at $0. Plan Bee lets you have truly pledgeless goals (as many as you like) that can start at and stay at $0 (and drop back down to $0 if previously at $5).
The fact that it takes a paragraph to clarify that means we’re doing something wrong…
My inclination is to solve this by yanking the “infinite freebees” perk from the premium plans. There are only ~50 people taking advantage of it and it’s pretty un-Beeminder-like. (But of course anyone who’s signed up on Plan Bee before we make changes will be grandfathered or better.)
In the meantime we’ve tried to disambiguate by referring to the Plan Bee perk on beeminder.com/premium as “fully pledgeless goals” and are inclined to try this “10 freebees for $1” idea since it’s easy and the feedback has been overwhelmingly positive. (Thanks again!)
But back to the motivatability scale, my idea was that if there are many of you lower on that scale then we should perhaps keep pursuing seemingly un-Beeminder-like ideas like unlimited pledgeless goals.
Here’s the final histogram of responses (also repeated from a recent daily beemail):
I propose for your consideration two reasons why you should take the
“lets change the rules objectionably, but grandfather people in” move
only under extreme circumstances.
It reflects badly on beeminder.
It imposes a permanent software engineering and relationship
constraint on beeminder.
I can elaborate on each if necessary. But I for one would much prefer
that you never grandfather anyone in. When you change the rules, change
them (and yes, that means having that hated “we get to change the rules”
clause). And as a corollary: change them (objectionably) only rarely and
carefully.
Beeminder has, at different times, expressed two views that seem to
conflict:
… the QS First principle, and in fact I view Beeminder as foremost
a Quantified Self tool
I think the infinite (subscription paid!) freebees are a very QS
feature, and hence very Beeminder-like also. People are telling you on
many threads that the QS aspects of BM are at least as important as the
sting. And for paying subscribers there is no conflict with revenue.
I personally love BM. It works for me for some important things, though
not yet for ugh-fields nor for work. But for me personally the “hurry
now, get grandfathered” calls are a turn off.
Excellent points, and I did have that in the back of my mind a bit when I added the “or better” after “will be grandfathered”. Like if it looks costly for the reasons you suggest (I’m not sure if it will be) then we could instead do a greater-than-100% refund on the most recent premium payment. We’d have to think it through and, mostly, ask the people affected. (Which, again, is not many.) Noted on the turn-off of anything with an “act now!” feel to it. Great point.
As for the seeming conflict between the QS First principle and fully pledgeless goals, my own feeling is that, yes, Beeminder is first/foremost a QS tool, but it’s still not exactly beeminding if you remove the pledge part. It’s just, y’know, minding. QS First is actually my own bias. Many of us view Beeminder’s graphs and data as purely a means for implementing flexible commitment contracts. [blog.beeminder.com/philip] Ie, QS Last.
But even QS First doesn’t mean QS Only, is my point. We gladly point people to alternatives like blog.beeminder.com/trackhack if they want QS Only. As you point out, it makes sense to give people that if they’re willing to pay for it, except, as you also point out, there might be a software engineering cost and relationship constraint, dealing with two very different kinds of users with different sets of rules and algorithms. (But it might be fine!)
I should also mention that we’ve avoided any grandfathering so far (other than prices for premium plans) and do want to continue to avoid it. So, yes, really glad you pointed this out.
Feedback still solicited on how bad it would be to drop unlimited pledgeless goals from Plan Bee, given that you could instead buy 10 initially pledgeless goals for $1. But I also think we may yet find a way to have the best of all worlds and make QS Only and QS First and QS Last people all happy in an elegant way. [note to self/inner-hexagon: pledge caps proposal may be part of this (might be worth a separate thread if people are curious)]
Thanks again for the highly constructive feedback on this stuff, Daniel!
I think that one blindspot that QSers might suffer from is that for most people, just one system is plenty complicated enough. There isn’t room in my life for beeminder and for iDoneThis etc.
Having the sting available in BM makes it the first QS tool attractive enough for me to try. The possibility of also using sting-free QS for some goals on the same interface doesn’t detract, so why remove it?
I find myself rambling in my argument. Actually, I just don’t know why you are considering this change. “not BMish” is not a reason, it is shorthand for something I still don’t understand.
Maybe conceptual integrity: “we only do stuff with a zing”? but zing=0 is just a valid special case (for paying subscribers). Have I missed some other explanation?
More generally, I would say: change is not all good. More speed is good. Less bugs is good. More (quality) integrations is good. New features… sometimes, but keep in mind the system basically works, and complexity is a bug. Removing existing features!? have a really good reason for that?
I think that one blindspot that QSers might suffer from is that for most
people, just one system is plenty complicated enough. There isn’t room
in my life for beeminder and for iDoneThis etc.
This.
I don’t think the conceptual confusion that exists is a good reason to get
shot of a valuable feature. Simply rename the premium version, as
suggested, to something like “totally pledge-free goals”. It’s (quite
reasonably) premium only, so it’s not like the population who might get
confused by this is very large.
On this topic, I’ve been thinking fairly hard about the post made a few
weeks ago which linked to The Black Arts of SaaS Pricing and what
exactly the best way of seeing the premium plans is.
The QS/free goals thing is essentially the only reason I have a premium
sub. This seems to be quite a fundamental point: your business model is not
dependent on subscription income, and (except for this case, where an
obvious benefit undercuts your business model) it’s not clear that they
offer many benefits beyond warm fuzzies of financially supporting a good
idea.
The second point is that your business is not like those discussed on the
Kalzumeus posts: you don’t sell premium subscriptions to corporates for
whom whatever subscription level is trivial and the glossy presentation is
all. You sell to individuals, for whom the “feel” of the service is
actually central to the basic offer and absolutely key to whatever added
value is offered by the premium sub. By making small chisels away at that
feel, whether it be by establishing a separate class of grandfathered
service or by making things “premium” which are actually helpful for the
delivery of the basic Beeminder service (which, let’s not forget, comes
with its own revenue stream) then my feeling is that you’re slightly
cannibalising your brand reputation.
In short, it seems to me that your approach should be closer to the Tarsnap
example that is derided (extraordinarily honest and straightforward, and
very much a person-to-person service) simply because your competitive
advantage rests on either your model not being copied (seems unlikely) or
on relatively geeky individuals (not large bureaucracies) deciding that
you’re absolutely the swell kind of people that they want to send cash to
on a regular basis to keep themselves honest. The premium subs (however
happy I am with them) are a bit of a sideshow.
I don’t for a moment think this is the only perspective on value, and
you’ll have spend orders-of-magnitude more time thinking about this than I
have, but it seemed important. Feel free to treat it as the unsolicited
advice it is!
Daniel and Tom, this continues to be brilliant feedback. I can allay most of your fears about grandfathering and yanking features. When hashing out this “10 freebees for $1” idea it started to seem too convoluted how Plan Bee is all about unlimited freebees but then you can get freebees dirt cheap but the Plan Bee freebees are freer than freebees because they can stay at $0 and OMGtoocomplicated!
But we hear you, throwing the QS Only people under the bus isn’t the right solution. So we won’t do that.
One thing I wanted to ask you more about, Tom: why would it feel to you like cannibalizing our brand if we made basic things premium? The extreme of that would be to make everything premium, like a limited free trial but then you have to pay to keep using Beeminder. A normal subscription service, in other words. The biggest argument against that is that we’ve made a big deal out of the “if you never go off track you’ll never pay anything” feature. If it’s just the seeming unfairness of having to pay a monthly fee when you’re already paying pledges we could fix that by applying all pledges as credits towards premium plans. (But we’re not thinking seriously about this at this point; mainly just wanted to hear more about possible brand cannibalization concerns.)
Thanks again for the help thinking this stuff through!
But we hear you, throwing the QS Only people under the bus isn’t the
right solution. So we won’t do that.
There are no QS only people, there are QS only features. Every person can
benefit from stingy goals and also from QS only goals. Only difference is
stingy goals statistically pay for themselves, but you have to charge for
QS ones.
Looks reasonable. Does add some complexity, but not much.
One thing I wanted to ask you more about, Tom: why would it feel to
you like cannibalizing our brand if we made basic things premium?
Here is a natural line of thought:
your paying customers are everyone that has given you a credit card, not just “premiums”.
hence moving features to premium = degrading service for paying customers.
if you are often degrading or even talking about degrading, then you will probably degrade in the future, forever.
maybe I should put my data, time, loyalty and money somewhere safer.
Hence cannibalizing your brand. Not worth it for omgtoocomplicated, IMO.
Going back to the basics, I think bm is the bees knees just as is. I am sorry if it isn’t generating as much income as you’d like, but hiking prices shrinks your potential market.
Daniel
The
extreme of that would be to make everything premium, like a limited
free trial but then you have to pay to keep using Beeminder. A normal
subscription service, in other words. The biggest argument against
that is that we’ve made a big deal out of the “if you never go off
track you’ll never pay anything” feature. If it’s just the seeming
unfairness of having to pay a monthly fee when you’re already paying
pledges we could fix that by applying all pledges as credits towards
premium plans. (But we’re not thinking seriously about this at this
point; mainly just wanted to hear more about possible brand
cannibalization concerns.
Am I reading your draft spec correctly in that the possibility of exponential goal growth is being removed from non-premium users?
I think this would take away quite a bit of the sting of Beeminder for me. I have two goals at $30 and I am terrified of derailing on them, not because I can’t pay $30 but because I can’t pay $270. If derailing the goal would leave the pledge at $30, I would probably derail more[1]. That, and the idea of paying money for the privilege of paying more money seems backwards.
That said, short-circuiting my goals to $270 doesn’t feel right either. Even though I am pretty sure $30 isn’t going to keep me on track forever, the longer I don’t screw it up the longer I get to stay at $30 and that feels a tiny bit good every day. This is all just ways of saying that I find the exponential schedule motivating.
[1] Yes, this may be more revenue for Beeminder short-term, but I’m not going to use the service if it isn’t able to get me to do things.
[2] Today I’m minding 19 goals and I’ve paid Beeminder a total of $60.
