Pledges that rise and fall automatically to maximise user-value

What if pledges were like little hot air balloons, whose level rose and fell within set limits, according to some simple algorithms, to improve your likelihood of keeping your commitments? Is this a good idea? What might those algorithms look like?

I was having that conversation in Support with a user, and with their permission here are a couple of exceprts, to get your thinking/arguing responses (and credit goes to @dreev to an earlier discussion about some of this stuff):

I started with:

as well as the existing pledge cap, have a “pledge ledge”. If your pledge ledge were, say, $10, then even if you start a goal at $0, over time (perhaps once a month) it would tick up to the next level, until it was sitting on the ledge. Similarly, if your goal had hit your goal cap of say $90, then - assuming no derails - over time it would decay back down to the ledge. So you would set the ledge at the “paying to break a commitment” level, whilst putting the cap a step or two above that, to provide a sharper incentive not to use the “pay to break” ability too often.

To which they said:

I have definitely wanted something like the pledge ledge, though I don’t really see why it’s necessary to have the pledge go up to an equilibrium point. Some kind of exponential payment schedule for breaches, but maybe the penalty decreases by $1/d until it settles back on $30 or so? Taking the “derailing it is nailing it” framing, you want a goal that’s hard enough and that you care about enough, so that you get stung occasionally but you’re generally succeeding, and the penalties get really stiff when you have sustained breaches.

What do you think - is this a good idea, or a terrible idea, and if it’s good, what might work best?


This is a good idea. It’s the pledge equivalent of what @mary first hacked into their personal dashboard and which @narthur implemented as the TaskRatchet autodialer.

I do this manually and inconsistently; after a period of keeping on track with a goal, I’ll reduce the pledge, sometimes in tandem with ratcheting the goal and reducing the safety buffer.

My suspicion is that those two things ought to go together; an auto-pledge (within user-comfortable bounds) and an auto-dialed slope (again, with some user control). Either one without the other seems unlikely to maximise awesomeness.


I have an n8n script that puts my pledges back to 5$ automatically.
It helped because I derail often, and I was neither happy calling off, nor paying 90$.
Why not putting the max pledge to 5$ right away?
Well, I sometimes end up with goals that I derail on repeat for weeks.
It helps to have exponential pledges to catch these and either archive the goals or pick myself up :slight_smile:

guide - automatically stepping down your goals with n8n