Idea: Automatically adjusted project goal

Project funding goals are kind of arbitrary since budgeting is hard. Projects might not know what goal works best for them.

So i had the idea that the funding goal is automatically set and adjusted by the system.

It might be that there is only one best option when and how to change the funding goal. We would need research on that. And if so, it would not make sense to let projects choose a worse goal.

So let’s see the dynamic:

We need to research which difference from currently-funded to funding-goal has the maximum crowdmatching effect (visitors are encouraged to become patrons). When the difference is very high (100$ funded but goal is 10.000$), contributing seems pointless. When the difference is very low (9.900$ funded from 10.000$), it’s very encouraging, but it will be 100% funded before the payout, so there is a time without the crowdmatching effect.

Also, we want projects to get funded, so the might not let the funded % drop below 50% (?).

So, how could it look like:

  1. A new project is created at the start of a month. Starting funding goal is 100$.
  2. it slowly get’s more and more patrons. they finally reach the goal after 3 months
  3. the system adjusts the goal to 200$. they reach the goal in 1 month
  4. the system adjusts the goal to 300$. they reach the goal in 1 week
  5. the system adjusts the goal in a way that it is likely that they reach 100% at the end of the month (when the money it payed out), but not more than 100%. in this case, the historic data shows a growth of say 100$ per week, so the goal is set to 600$. at the end of the month 700$ of 600$ goal is funded. so 100% of the pledge is transferred. the goal is not changed when the 600$ goal is reached, but after the payout (so it don’t drop below 100%)

Now assume a stagnating project

  1. the funding goal is 1000$ and it is 900$ funded. it looses 50$ more funding in the month. so the system adjusts the goal down to 900$.

so while a project loses more and more funding, they still get say more than 80% of the pledges and new visitors have high motivation to become patrons, since it looks like it will reach 100% soon. so the project has a good chance the funding grows again, when they do good work.

So we need an algorithm, that sets a new funding goal after the payout which tries to get the project 100% funded at the next payout, but maximize getting more patrons in the mean time by letting the funded% not drop below 50%. It should also not get more than 110% funded.

We see that the algorithm is successful, when a growing project mostly get’s 100% of the pledges (that is the goal of the project and the patrons), but also get most of the potential new patrons (we have to track the conversion rate of visitors). A stagnating project still get’s 80% or more.

So the algorithm is optimizing funded% at payout AND new patrons. Simple min-max strategy. It’s probably simple machine-learning forecasting and not even deep learning, but maybe that is usefull too. Data is the historic projects funding, also from all projects and new patron conversion rate of all projects and from the project itself weighted more. Maybe a strategy work for one project, but not for another.

With this, we would offer very simple and efficient funding, since the project just has to create account, setup payout, create project and continue working on the project. We acquire the maximum potential funding for them.

1 Appreciation

I’m going to need to sit with this for a while, but here’s a few initial impressions:

  • I appreciate the simplicity that it offers incoming projects
  • It may be difficult to explain to patrons, and for that matter, projects
  • What would it look like if projects still had the option to select an opening goal?
  • We should really generate a variety of model plots…
2 Appreciations

It may be difficult to explain to patrons, and for that matter, projects

yes

but it’s not that different from a szenario where a project is really dedicated to optimize the goal and change it every month. we can automate that for every project. computers are very good at such tasks : )

What would it look like if projects still had the option to select an opening goal?

  1. project sets start goal of 10.000$
  2. no one becomes patron, because it is pointless
  3. at the end of the month, the system sets it to 100$

so we can let the system set a useful goal directly and not waste a month and leave people with the impression, that funding this project is pointless. such a first impression is hard to change

That’s not how I see it. Contributing seems costless to me. It’s essentially a bit of extra effort version of an upvote. I get to put-my-money-where-my-mouth-is and say “I’m in!”. And it costs me no money really, until the crowd grows which is then wonderful.

I feel the opposite. At that point, I think, “well, some other people will come along and hit the goal without me probably, so I can go use my limited resources elsewhere”. This is where I have tensions around a threshold and want to see stretch goals. I’m only encouraged when I think I can actually get those thousands of others to donate more than they would without me.

Budget control?? Matching??

So, the point of having a target at all within the mechanism itself is to give patrons a budget-control point. They are stating willingness to match others to get to that point, and they know what their max is. How does that work with your proposal?

If I, as a patron, am either (A) spending the bulk of my whole budget most of the time or (B) am getting my budget manipulated by the system without my opt-in, this seems unacceptable. The whole point is that I want to offer my willingness to give more if I’m part of a larger crowd and I need to know that ultimately, I have some control over how much money I’m making available to the system.

I don’t see immediately how these things work within your idea.

2 Appreciations

If I understand you right, your goal is to maximize the monthly payout.
Wouldn’t that mean to ultimately equate the goal with the current pledges – basically rendering the whole mechanism absurd?

2 Appreciations

Yes. Or you could say “fully utilizing the financing potential of the community”.

In best case, you get the full pledge of contributors, but only when the project got X new patrons. The system predicts a realistic X.

That idea might only work well with unlimited growth. I haven’t thought about it that deeply myself.

Think about a project that gets 2 new patrons a month. The system would set the new goal to + 2 x average pledge. That might not makes sense. It could have 100$ steps…

No, it equates the goal with the potential pledges at payout time. The mechanism is still there and offers visitors, become a patron now, so everyone matches you and we can reach our goal THIS MONTH.

The goal is just there to motivate visitors to become patrons! It has no meaning to the project. We offer the project, we get you as much funding as possible from your community, better than Patreon.

The goal might get reached some days before payout, then you are right. There is no matching. But that can also happen in the case that the project sets the goal and stay that way until they change it. It would probably be faster when our system does that.

So you think you can just upvote any project you use or like? It will cost you when they actually reach the goal and you might not remember that you “upvoted” 100 projects a year ago and have to pay 1000$. So you really should think about the cost that can happen and only “upvote” what you can afford.

I think people will think that a 1% funded project is not successful and not worth investing in. When we keep it always at 70-90% funded, new visitors will think it is successful and worth investing in.

That might be possible. But your resources have a higher impact here than on a lesser funded project!

When a project has set the goal and announced that they will be able to afford X and you want that to happen, you have a motivation to contribute. That might be the strongest argument against this idea.

Crowdmatching still happens, since the goal is not reached.

I would disagree. Patrons have no control over the crowd size, only over their max.

You are right that we loose some kind of control because the system changes the goal to some amount that makes sense to the system.

My assumption is

the patrons want to give their max! WE hold them back because we claim that our mechanism can bring in more patrons that way.

see patreon: people giving an amount and just accept that most of the community don’t. that would be their max in our system

A: again, i assume patrons want that
B: i don’t see how the system manipulates anything. patrons set their max and thay pay it or less, never more

it’s kind of the same here. you give more when the goal is reached

only that the goal is reachable in a month

when the goal is not reached many months, the mechanism holds back potential funding. there has to be a good reason for that!

we claim that we can acquire more patrons that way, BUT what if that also works as well with this idea or even better?

This idea might be too wild and we should discuss the dynamics of changing the goal without a computer doing it first.

I create yet another thread… Idea: Projects have multiple goals

[mod edit: I added the link to the thread. Please note in the future the useful reply-as-linked-topic feature here, see the animation at About the Forum meta etc category ]

1 Appreciation

I could have been more clear. The upvoting in this case is not for “this is a good project” it’s for “this project deserves more funding, and I’m willing to help”. Say there’s a poll that says “Do you think this project deserves more funding and you’d be willing to help?” Pledging in crowdmatching when the crowd is small and starting out is just barely more hassle than saying “yes” to that poll. It’s a commitment to actually be willing to help, not just say that I would.

This is why our current model emphasizes a system-wide budget. There’s still some value to that. And we could potentially offer some automated system-wide-budget-cap anyway along with the project-goal style crowdmatching. There are various ways that could work. It could be that I set for myself a buffer I want to have before I add any pledges. It isn’t necessary per se, as someone could choose for themselves manually to manage their budget. We should certainly display in each user dashboard the sum of all their pledges, even though they won’t be all maxed out in practice.

You’re not investing in it if you are only giving at 1% of your max. That’s the whole point. And while some people act like you describe, we know from prototyping and discussion with lots of people that this is not how everyone thinks. The project is not a failure because (A) the project is successful enough if you can already use and appreciate it as is (we’re focused on projects that have some existing published project already) and (B) it depends how long it’s been trying to get funding. Nobody says that a Kickstarter project has failed because it only has one pledge in the first hour it’s been open. A project that reaches only 1% of goal in the first month but (A) has an ambitious goal and (B) shows growth curve, nobody will look at that and just say it’s a failure.

This is basically dishonest. You’re setting up a situation where all projects are artificially adjusted to look like success. A scheme where every project succeeds is not trustworthy. In the long run, we expect that some projects won’t actually have much support and that’s okay. Failure of sorts has to be possible or the whole thing is a farce.

I’m thinking of the practical expectations. When the project is 99% funded, if over the next week it’s going to hit 100% whether or not I pledge, then I’m not actually getting matched. Matching has to extent past 100% into a stretch goal for the matching premise to fully hold. Otherwise, if I help get to 100% and that results in someone else deciding not to pledge, then I’m simply shouldering the burden instead of them. The freerider situation remains.

But this is arguing against itself. I’m imagining you are seeing things from your lens as a passionate FLO person who has already made donations in the past. YOU want to give your max. But if it were true that “patrons want to give their max” overall, then there’s no “more patrons” to bring in! Who would they be? They’d be in already if their goal is to give their max! The whole point is that potential patrons mostly do not pledge at all and many who would be open to giving do not give. Current donation systems allow anyone to give their max if that’s what they want.

In short: your claims are incompatible. We cannot both believe that there’s a huge latent pool of hesitant potential patrons and that everyone’s main goal is giving their max.

What I’m saying is that most people have a goal of freeriding and getting all the value without cost, and a large portion (but not 100%) would be willing to give something to help the value exist, but they’d still rather give less than give more. I would rather donate just $1 to Inkscape or even just nothing and already have it be everything I wish it could be. I have no goal of spending $10 or $100 as an end in itself. I’m just willing to give up $10 if that’s part of actually realizing the value I want out of it.

We could run prototypes and tests of dozens of variations if we had the resources. I am skeptical enough of your idea to not support investing much in researching it. But you could certainly talk to a handful of acquaintances and try to get some sense of whether it would actually get traction. In the end, we have to do our best to find the best-fit ideas based on existing knowledge and research in order to figure out what models to even invest time in discussing.

I see this as counterproductive. The primary purpose of having a project goal at all is not just to play some game with monthly levels as part of an algorithm, it’s to express something concrete about the vision the project has. We want projects to express what they could do with more funding. Goals that are really modest and constantly iterating upwards slightly do not offer that sort of vision.

The other purpose of a goal is to help patrons set their budget based on how much they are willing to contribute to that vision. So, if patrons just see an arbitrary goal of a modest amount, this is backwards. If I see a project seeking only $200, I don’t think “I could give $20 of that”, I think “meh, some other people could help get them there, I don’t really want to be one of the few self-sacrificing people who help them get to $200 while the rest of the thousands or millions of users freeride”.

When the goal is ambitious, say $20,000, then I more readily say “if my $20 is part of getting this vision to be reality, I’m in. And if this dream fails, at least I’m not out my $20, and maybe it’s at least something positive if they get to $2,000, and okay enough then if I put in $2 of that.”

4 Appreciations

it’s not failed (yet), but it don’t look good

the other way around, a project that is 100% funded after the first hour is a big success

not a failure. my point is, it looks worse than a project that already has many backers (people who stand behind the project). and the number of backers in relation to project size, simply shown by funded%

when firefox had 100 patrons, it would look unsucessfull, because you know many people use it, but only few of them actually back it (is there a good reason not to? why should i when they don’t?)

a mastodon instance with 100 users and 100 patrons is a huge sucess, because ALL (100%) contribute financially. this would be the best case and new users would be encouraged to become patrons too

well said. i agree

So, freerider situation remains when goal is 100% reached. What can we do about that?

We can thange the goal when that happens, but the project will never get 100% of the pledge.

i think patrons want to give that amount because they choose it (for whatever reason). as i said, i see that on patreon. they hope it will have an impact

with crowdmatching, they only give when it actually has this impact

and “this impact” the patrons expect is what the project communicates as goal. so this communication really is essential

yes. i was thinking about existing patrons. they give an amount, so i assume they want to

that totally makes sense

i actually also just want to give when it has an impact. sadly, it often has not

but that would give the project more actual funding. we should be aware of that. such thoughts are what i think @mray mean with gaming the system

and my idea was, if that is possible, why not make it available to all projects

in our system, projects with bigger goals seem to get punished. to get much money quickly, just set a low goal like 100$

i ignored that view and just played with numbers to see what effect it has

1 Appreciation

Celebrate! We achieved a fully-funded public good! It would be nice to have incentives to reduce freeriding up to the point that any stretch goals and a buffer exists. But freeriding is not itself a problem unless someone is dogmatic about the principle of it (in which case, they basically oppose public goods).

I want more of my neighbors to help clean up litter (something I do actively). I want less people littering. I don’t actually care that every single neighbor pick up litter. If it takes 10% of us to keep the neighborhood clean and none of us feel very burdened or burning-out being part of that 10%, that’s fine. We have a reliably clean neighborhood. I have higher priorities than caring about the unfairness of the 90% of the neighborhood freeriding.

A tiny portion of people who are passionate just give all they can and hope for impact. The vast majority of even those who give on Patreon only do it to (A) get the [usually closed, proprietary club-good] bonuses and/or (B) feel even better about their own view of their generosity. A few at the high-end do have a significant impact, but they are likely wealthier anyway. Near-zero ever give to the point where it hurts financially. Nearly all of them actually could give more without it hurting but mostly just give enough that they can feel good about themselves.

With crowdmatching, the emphasis becomes more pro-social and more collective-action. It feels more like joining a team, a movement. And overall, people will still set their max based on what feels like “this is a fair level for me to be doing my part” rather than what actually is the limit of what they could afford.

Yes, this is a real concern. But this is why the vision matters. I may think $5 for a project is fine, but when I realize the gap between status quo and the potential is wider, I’m more willing to say “oh, well, if this were world-changing impact, I mean, I’d be okay with $50 or more to reach that dream”.

In the end, I think the balance involves having multiple goals indeed. But they should be social and vision based. They should be about what the project can accomplish. I imagine a normal of something like $1000 for more reliable sustaining support, $5,000 for real robust improvements, $50,000 for world-changing etc. So 3 goal levels seems good to me to start out. People can pledge so that they are really in on the make-this-succeed-now modest level and also have a big long-term vision. We could have a system that has reasonable interactions for how the pledges work across these levels. The levels would remain tied to project goals for real progress, not just arbitrary dollar amounts.

I know @msiep and @smichel17 have thought a bit about the options for a handful of tiers like this and how they might work. I support that direction very much.

1 Appreciation

@wolftune said most of the things I would have, but I have a few to add:

Well, it depends. I would not consider it a big success if the goal were $1.
Then it would be a small success :upside_down_face:

This difference is exactly why different projects need different goals, instead of a constantly-adjusting one. 100 patrons giving normal amounts is enough to make a difference in funding a small mastodon instance. It is not enough to fund Firefox.

For mastodon, the goal should be small enough that 100 patrons can reach it — otherwise we’re leaving money on the table. On the other hand, it’s a bad thing if 100 patrons gets Firefox near its goal, because those patrons will be wasting their money.

Say I am considering becoming the 101st patron of firefox, and I can afford to give $30/month. I am NOT willing to give that yet when firefox is so far from their goal. As you said,

why should i when they don’t?

…and this is really the heart of crowdmatching. It gives me a way to express “I am willing to give $30/month, but only when others are giving with me and it will make a difference.” If we automatically adjust the goal so that it’s always near 70%, I am no longer able to express the “but only” part… except by not joining the crowd.


Reframing the whole thing: I think it is a much better look to have 100 patrons of firefox, each giving 0.1% of their maximum (basically nothing), than to have 100 patrons of firefox, each giving 70% of their maximum, and still accomplishing basically nothing.

( This is a feature, not a bug :slight_smile: )

3 Appreciations

Wouldn’t you agree that every end of a month the potential pledges are 100% of existing pledges? …most sensible time to do optimizations would be right before the payout, I suppose.

…not sure whether I’m beating a dead horse here (didn’t read the whole thread)