As a point of reference for discussions in the meeting I want to post a rough sketch of a “mechanism 1.1” that is based on matching people, not money:
Projects set a monthly goal: Number of Patrons
Patrons choose among levels of pledging (e.g. 1€, 3€, 5€, 10$ )
Goal achievement in % equals the % of each pledge actually becoming a donation, surpassing the goal starts to lessen every patrons burden proportionally again.
(minor detail: In favor of the project, surpassing the goal would probably count high level pledges first, only “redistributing” lower level pledges among patrons)
rough comparing visualisation (actually valid frames would snap before/after a patron):
I think this is a good place to discuss this idea vs $-based matching, so I moved two related posts here (below) from other threads. (In case anyone is wondering why they don’t respond to @mray’s post directly)
I feel strongly that dollars is the only option that combines sufficient simplicity and flexibility. I think the focus on number of patrons needs to be approached as a goal for how we communicate and frame crowdmatching, but should not be part of the math of the mechanism. (And certainly projects should not choose what type of goal to have.)
@mray and i discussed crowd-size vs money goal last week. i try to summarize it, but feel free to add your opinion
@mray argues, that both have the same outcome at 0% and 100% of the goal, because we can’t control how many people actually pledge and what they decide to give, so we assume it’s the same in both cases (or no significant difference)
i think that is not correct. we have an influence there. when we present our funding mechanism as successful and the project presents their funding goal as desirable, people will more likely pledge
and while it is nice when every patron has the same weight in achieving the goal (like in crowd-size), i think it can motivate people to give more when that has a direct effect on reaching the goal (money goal)
we want the project to be independent from single patrons, but that can be solved with limiting the amount one can pledge. i think that could in any case be 100$ (losing that amount has no big effect on projects. you don’t have to fire an employee) AND also say 10% or 1% (when the whole goal is to collect 100$, no single patron should pledge 100$, but only X% of that)
i think there should always be the option to pledge 1$ (so no one is excluded) and when 100$ is the max, that has a 100x effect on reaching a money goal. would that be OK or is it unfair towards the people that can’t pledge so much? poor people can feel powerless and we might not want to have that also in our project… maybe 50x is less worse? but maybe we dislike the idea that one person is X times more important than another ones… but when we don’t show how much the other pledge, people can only calculate the average, so i think it is not a real problem.
also, when a project communicates how much money they need to achieve what, it is a way more desirable goal than “we want to have 2000 patrons” (for whatever reason). in the end, it’s about funding real work with real money!
so, while i believe that you can build a system with a crowd-size goal, i don’t see how it can be superior to a money goal.
in today’s meeting, it was also said that a money goal is more similar to other popular platforms like patreon, so people can understand it more easily
(that ended up very opinionated from my side, so feel free to add other perspectives)
I am skeptical of how practical it is to ask projects to choose a goal in terms of crowd size. However, currently I think that patron-based matching is theoretically superior. Solidarity and other arguments not needed; it boils down to this:
This reminded me of the whole premise behind Snowdrift.coop, the reason we think Crowdmatching will work: that the vast majority of people who are willing to donate, currently don’t. Crowdmatching isn’t about convincing the people who already give to up their levels, it’s about changing the status quo by giving people a reason to donate.
Someone who’s been a minimum-level patron for a while may decide to raise their pledge, or support other projects, later (the next time they visit the site, which they’re likely to do, since they’re a patron now). Someone who closes the tab may never think about Snowdrift.coop again. So, let’s optimize for getting people over the initial hurdle, by making the first step as easy as possible and lining up the incentives to encourage becoming a patron, at all.
With dollar-based matching, you also need to weigh the benefit of a higher pledge (and others matching you more) against the higher cost. With patron-based matching, the decision is broken into two simpler ones: whether to join the crowd, and how much to give. Once I’ve decided that I’m in — I want the matching — it’s easy to punt on the second decision and just give the minimum, unless I really care about the project.
I’m imagining a something like this (details not important, just trying to get across the main idea):
This highlights something really important: while related, the goals a project sets (and how they are communicated to patrons) are distinct from our choice of crowdmatching mechanism.Any goal can be presented as both a dollar and patron amount (only one used for matching).
How
When matching based on a dollar goal, divide by the highest pledge level we allow to get the patron goal (and round to 1, if there is no limit ).
When matching based on a patron goal, multiply by the minimum allowed/possible to get the dollar goal.
The current system can be expressed equally well either way, since the minimum and maximum are the same.
It’s totally valid to say, “we’ve reached our dollar goal, but not our patron goal, so matching is still active.”[1]
This is where we come to multiple goals.[2] With a single goal, setting an exact target is very important, because that’s when matching turns off. But with multiple goals, it’s okay if things are a little fuzzier, because if you underestimate how much you’ll need for your first goal, your second goal can pick up the slack.
I really like these particular tiers as a framing device. They’re quite distinct, and that makes it easy for me as a patron to consider how much I’m willing to give at each tier.
We talked a lot in today’s meeting (breakout room) about the inherent fuzziness in any sort of goal-setting. More so with trying to pick a patron goal, but even with dollar amounts — our goal is really “hire one person to work on Snowdrift full time”; how much money that requires depends on who we hire.
That’s our “sustaining” tier. But let’s jump up to “world-changing”. That’s so far away, it’s kind of hard to imagine what kind of resources we’d need. In this case, I think it’s probably easier to estimate the largest-sized crowd who’d be willing to donate to Snowdrift — off-hand, I’d ballpark ~100 000 people. (And, of course, “improving” is somewhere in between).
Coming full circle: especially at lower tiers, I think it makes sense to ask projects for a dollar goal. I can’t imagine us, at our current stage, defining a patron goal. But that doesn’t stop us from calculating a patron goal, based on some reasonable minimum pledge, and matching based on that. My main concern is whether it’s possible to do this such that the numbers always work out to be reasonable — if the minimum is $1, and our goal is $50 000 per year (~$4150/mo)… 4000 patrons seems like a lot for a “sustaining” goal. But then again, we already have 127 patrons without any publicity and with our website in its current state, so maybe my assumptions are off because I’m used to the status quo.
We could also do something more complicated, where both goals matter, and matching is based on whichever one we are further from achieving. Not proposing that. ↩︎
But not the ideas I’ve been alluding to, sorry. They’re about how to do calculations with multiple goals, and this post is already quite long. ↩︎
I have a visual conceptualization of two slider bars, one for dollars and one for patrons. When a project initially sets up their page, the sliders are linked, but there’s a checkbox to unlink them. It gives immediate feedback regarding their connection and allows the project to select the numbers based on what they value, not what we value.
Interested in further discussion of multiple tiers, but don’t think it’s a launch feature.
The significance of a single “I’m in!” pledge button default seems really good to me.
People absolutely must understand what they are agreeing to when they pledge, of course. So, it does seem to me that focusing on the idea of the crowd and the other patrons is good, but when we’re framing what the crowd is accomplishing for the project (i.e. what the public goods are that we hope to enjoy), then it makes more sense for many reasons to have that be dollars.
So, I’d say we want the focus on patrons and the crowd when we describe the matching, and we want the focus on the dollars when we describe the goals for the projects.
I think dollar-goals can be an option but still prefer the crowd-goals approach. In terms of flexibility, simplicity and doing calculations I can’t see dollar-goals being superior or crowd-goals unfit. I do see notable differences, but those don’t make me discard either approach.
As the dollar-vs.-crowd question seems to be the most important and controversial right now, I’d like to dig a bit deeper:
How would you define “sufficient simplicity” and “flexibility” in that context?
Why do those properties matter in your eyes?
What makes you skeptical about crowd-size being part of the mechanisms math?
It does not work that way. Comparing two things with the assumption that both start with the same premise you can’t unilaterally change the premise after the fact. By definition you don’t have an influence on something you assume from the start.
A brief summary would be: it puts people before money. This isn’t just a strong message, but gives a framing to the entire project (which is why I think this question is so important).
We’re about gathering a crowd, and the fact that we ask for LESS!! than the maximum clearly shows that money isn’t the very first thing we care about.
While I agree on how crowd-goals speak to people more personally (not as a source of money) I think you can’t say this:
Whether – or how much to give, are the two questions you face either way.
You seem to point out another concern here that affects both approaches identically: setting a realistic goal turns out to be hard. Especially if you’re pressured to do a good estimate.
…A discussion on multiple goals and their possible automation to optimize them has its own set of severe problems and is a bit off-topic …
project goal stated as a rough output for the public goods and defined as a dollar threshold
limit on the percent of the threshold any one patron may contribute
thus the crowdmatching emphasizes the crowd size needed to reach the dollar goal
it is thus a crowd size ranging the number needed were everyone at the minimum pledge to the size where everyone is at the maximum
I definitely want to see stretch goals for past 1.1, so even though we won’t have them at first, I want to be sure that 1.1 version of crowdmatching will be amenable to adding stretch goals.
Well, broadly, I think the mechanism must allow projects to express a vision of what they will be able to accomplish if the goal is reached. While, a one-time campaign might say “we will get X Y Z features built”, our sustaining, ongoing situation, it needs to be more like “we will build this size of a team able to develop at this sort of pace, and accomplish these sorts of goals”.
So, what does that mean for the mechanism?
Given having different pledge levels available, there’s either (A) a range of crowd size that could reach a $ goal or (B) a range of $income levels that might come from a given crowd size.
I was initially hesitant about the crowd-size goal because I suspect the income range could be too wide. But maybe that’s okay? It’s an estimate anyway.
So, some random numbers to throw out:
Crowd size goal of 5,000 patrons
Minimum income could be $5,000 and max could be $50,000 (if it’s possible for everyone to pledge $1 or $10)
That range is way too wide for the project to give a coherent vision of the state when they have that range of monthly income.
I’m okay with a crowd size goal and expression of the range of income that the crowd of that size might deliver. But the range must somehow be small enough that it works as an estimate for the same overall project vision.
@mray I don’t think I followed the spread-the-burden mechanism you were suggesting. Such a thing could result in the narrowing of the income-range that the crowd would provide. Can you try again to describe that?
My biggest concern about spread-the-burden mechanisms is how they would interact with stretch goals. I’m inclined to see spread-the-burden as not part of 1.1.
That isn’t the number that they see though. As the crowd grows, they have a darned good idea of approximately how much they will be earning. Since they will be able to change their goal (somehow) they can decide if the crowd size target is worth modifying based on the makeup of their patrons.
Side thought: What if crowd size goal were able to be increased at any time, but the patron’s payments per month froze until the crowd caught back up?
@mray and I discussed things more in live chat, especially around the ways I was mixing the issues.
He really opposes mixed signals. If we go with crowd-based-goal mechanism, we should also be focusing on the crowd in the presentation and communication etc. That makes the decision more stark compared to my ideas of describing the vision one way and the mechanism another way.
And I see his point. This binary decision isn’t just about the mechanism with the flexibility to moderate it in presentation. Put another way: we shouldn’t use a crowd-based mechanism and then emphasize dollars in terms of the way projects express their visions. If we go with crowd-based mechanism, it’s because we want to frame everything that way. Money will obviously be part of it, but we don’t want anyone confused about how the mechanism works.
Crowd-based mechanism is really more radical. It’s the framing that focuses more on collective action, on cooperation. It sticks to the idea I already intuitively prefer: we don’t really know exactly what the potential of anything is, it’s all speculative. There are so many factors about how money may get used, so many ways to be wasteful or frugal etc etc. We want projects to have a business plan, but like our current open-ended model there really isn’t some true threshold, it’s arbitrary largely.
Going with crowd-based is a stronger rejection of the money-is-everything attitude we see in the world around us. It embraces the more political-activist framing and community. It’s more clear that we’re building a political movement rather than making just some financial tool.
I think we should take the time to fully embrace this direction in our discussion. What are the ramifications? Let’s explore some realistic modeling.
The meaning of crowd-based goals
How rough should project goals be?
I have resisted thresholds because they seem so arbitrary. Most budget estimates end up wrong. We haven’t laid out what exact amounts of dollars would do what even for us!
When @msiep brought up dollar-thresholds, I first accepted this as the push to do the (uncomfortable) work of figuring out our actual budget needs and estimates. And if we can get such goals that are really concrete and meaningful, I see value there.
Yet after thinking about this today, I’m liking the crowd-based goals and accepting more the arbitrariness. It’s not entirely arbitrary and it has a rough correlation with the actual income. Perhaps it’s better to keep the goals communicated as rough estimates and not have the pretense that we can really know in advance what the results of the funding will get us.
The whole point is that we aren’t putting a simple one-time price on a one-time goal like Kickstarter. We’re talking about sustaining and growing long-term. It’s even harder to estimate how things will play out. What we want is to let patrons review project progress over time and decide with real-world results whether the projects are making good use of the funds.
With the crowd approach, we’re asking projects to focus on crowd-building as the emphasis, on social validation more than just money. The pledges are more real in terms of put-your-money-where-your-mouth-is, and it’s real income for the project, but it’s about getting everyone to have skin-in-the-game and building resiliency and collective action.
I think I’m now with @mray and favoring crowd-based mechanism more solidly.
goal is met 50% (or anything under 100%): project gets 50% of pledged donations
goal is met 200% (or anything over 100%): project gets 50% of pledged donations
This gives incentive for the project to not set a too low or too high goal.
Set the goal too high: patrons withhold large parts of donations
Set the goal too low: patrons benefit from reaching the goal
Clarification:
after hitting the goal projects would’t start getting less money. Patrons just can afford paying less while keeping the the project where it is.
).