This is really good that we are discussing this, because it needs to be discussed. I think we all agree that it is best if we start out simple because it will be easier to explain and much more transparent.
I think the solutions that have been thought are really good and we will need to decide on something simple to start and then adjust to more complexity as the site gains more users. I would just like to suggest another solution: 1 pledge = minimum + minimum_matched (+ any extra the user wishes to donate) That means that patrons can easily donate as much as they want. The extra that they donate is not matched and does not affect pledging. A pledge is a pledge. Every pledge is matched and it doesn't create a problem if people want to be really generous. I don't believe that people who want to donate more than minimum are driven by the desire of being matched by 100s or 1000s of people. I believe that 99% of potential double-pledging will be from people receiving the money. (and I don't believe there are going to be many people who are greedy AND working on FLO projects at the same time). However, all of this is based on speculation so we really need to do testing. Maybe patrons need to have the extra incentive of having more matching or maybe not, but this is simple to implement and allows us to more easily incorporate the formula [/w/en/formula] as we go. I don't think we should design so much of the system based on preventing double-pledging. There are a lot of ways we can prevent it. But I really don't believe people will be motivated to turn their $6 into $18 for someone else when they can already turn their $3 into someone else's $9. Maybe I'm completely wrong, but I think we need to test it against the real world and then revise. I'm not saying that my idea is the right way to go (I have no evidence), but I am strongly suggesting that we choose something simple to begin with and then make it more complex later. :) Jacob On 10/20/2015 7:45 AM, Aaron Wolf wrote: > > > On 10/19/2015 03:40 PM, mray wrote: >> >> >> On 19.10.2015 17:47, Stephen Michel wrote: >>> >>> >>> On Mon, Oct 19, 2015 at 11:27 AM, Aaron Wolf <aa...@snowdrift.coop> wrote: >>>> >>>> >>>> On 10/19/2015 08:20 AM, Stephen Michel wrote: >>>>> In short, I don't believe we actually need any change to the mechanism; >>>>> we just need to lower the minimum and encourage donation at >>>>> above-minimum levels. >>>>> >>>>> We should do this by keeping in mind that *the average user will >>>>> tend to >>>>> stick with the defaults.* Therefore, if we set the recommended pledge >>>>> level above the minimum, so long as that pledge level is reasonable >>>>> (ie, >>>>> easily within the user's budget), they will stick with that donation >>>>> level. I propose the following. Note: numbers are rather arbitrary, I >>>>> just wanted to give a concrete example/idea. >>>>> >>>>> let n refer to the number of users. >>>>> >>>>> - Lower the minimum contribution to $1 per 5000 users. >>>> >>>> There's no basis for you to speculate that this lower minimum makes any >>>> sense. These types of changes are only sensible once we can operate and >>>> see how the numbers play out. Our current baseline is as good an >>>> appropriate guess and easier to calculate and explain. >>>> >>>> I think you need to read https://snowdrift.coop/p/snowdrift/w/en/limits >>> >>> I have read this. >>> >>>>> - For small n (< 100), the recommended contribution is $1 per 1000 >>>>> users. >>>>> - For n <= 100, the recommended contribution is the average of other >>>>> users' contribution. >>>> >>>> We don't want to recommend people counteracting the network effect. That >>>> would mean a message to others that says "if you join, others will >>>> adjust their pledge downward and actually *not* match you really". >>> >>> There's very probably some phrasing improvements. However: >>> >>> 1. When you join, others will match you at the level they have selected, >>> no matter what. The messaging should be: >>> - "If you join, current patrons will donate $X more." >>> - This is a simple concept which everyone gets. >>> - "Future patrons will match you at a level they choose. We'll >>> recommend they match you 1:1, so if you donate more/less, we'll >>> recommend they match YOU more/less respectively. >>> - This is slightly more complicated. Probably, this puts the idea >>> outside the scope of our MVP. This is fine. >>> >>>>> - This is presented to the user as "match other users 1:1" >>>>> - The user has an option to match at a different rate, but it's not >>>>> highlighted visually. >>>>> - If a user does opt to change their rate, the following message is >>>>> displayed: >>>>> - "This will [increase/decrease] the recommended donation[!/.]" >>>>> >>>>> Hopefully this allows for all of the following: >>>>> - A social incentive to donate more (increase the recommended >>>>> donation). >>>>> - A way to donate less with a reasonable social "penalty." >>>>> - if there's no "penalty," people may try to calculate the "best >>>>> deal" >>>>> of matching, ie, always donate the minimum. >>>>> - if there's too much "penalty," it may dissuade people who actually >>>>> can't afford it from donating. >>>>> - An elegant way to handle higher and lower contribution levels (ie, >>>>> adds little complexity). >>>>> - An intuitive way to present higher and lower donation levels to >>>>> users. >>>>> >>>>> Thoughts? >>>> >>>> All these goals are captured in our initial formula: >>>> https://snowdrift.coop/p/snowdrift/w/en/formula >>>> It has all the right properties to encourage larger pledges, discourage >>>> reducing your pledge, *allow* reducing your pledge… and we even >>>> originally started with a minimum that was a tenth the size of the >>>> current proposed minimum. So your thinking is exactly where we started >>>> with all this. >>>> >>>> The problem is that all this just leads to too much complexity, too much >>>> to explain, too many qualifications over the plain pledge concept, and >>>> so we really need to focus on launching without all this for now. The >>>> explanation of it all is just too cumbersome. The principles would be >>>> ideal to have, but we can't make it work practically. >>> >>> This is where I want to stress the difference here: ***this is NOT >>> actually a change of formula!*** For any individual patron the formula >>> is still, "I'm pledging to donate X per patron." >>> >>> It's a change in the level of donation that we recommend. Actually, how >>> we decide on that level of recommendation can be *completely* hidden >>> from the user. >>> >>> ~~Most~~ All of this is simply a change in how we present information to >>> the user. >>> >>> The point, is, I believe this, IFF we do the messaging right, is a way >>> to get back some of the benefits of the original pledge mechanism, >>> without a substantial increase to the complexity of the system, >>> particularly from the user's point of view. >>> >>> Most users will only ever see a button like: >>> >>> /-------------------------------------\ >>> | **PLEDGE** | >>> | (recommended donation)" | >>> \-------------------------------------/ >>> (change donation level) >>> >>> ~Stephen >>> >> >> This *still* is too complex - especially if it is not transparent how we >> get to our recommendation that we hope most people will just fall back to. >> >> I also have an issue with turning nobs *at all* in a system that should >> be a big knob that turns itself. >> If i understood you right it boils down to "smartly" adjusting a vital >> property of the user preferences that control the flow of money *on top* >> of our matching system. >> > > To be clear, we deprecated the old formula *not* because it was too hard > too implement (although it certainly did add complexity there) but > because it was too opaque and confusing to *explain* and present to > users. In that regard, the idea of some sort of variable "suggested" > pledge that fluctuates is just as bad, and it's just way further leap > than we can manage when people are already dealing with getting > comfortable with the concept of network matching at all. > > We *could* have some levels of suggested pledges based on historical > data and averages and other stuff later. I think it's a mistake to try > to get any of that now or really even discuss it much. > > I do even see the argument that the most clear approach is a simple > universal pledge with no settings at all. I still believe there's > substantial value for all sorts of reasons to the general ability to > pledge at higher or lower levels though. I was wary about people setting > up duplicate users to get extra matching, and that's the *main* reason > we had the old formula do tapering-off matching of extra "shares". > > I still think there's something to getting duplicate pledges be > something we endorse instead of something that games the system with > duplicate users, but it's too hard to manage easily because explaining > to people that there are X patrons by Y of them count double and so on, > it's just weird. > > But we're hurting our goal of funding if we tell wealth patrons that > they have to pledge only the minimum. We welcome wealthy patrons to > pledge more, which means both donating more and being a greater > invitation for others to join (because it means greater matching). So, I > really insist that our launch retain that basic function. The ability to > pledge at different levels is also aligned with *all* the donation > systems and crowdfunding systems people know already, so it will feel > familiar. > > And as much as we should be aware that we don't get multiple shots and > controlled experiments to run multiple ways, we *can* pivot still. We're > not irreversibly locked to our launch decisions. I'm not using the > ability to pivot as an excuse to not take our best aim initially. > _______________________________________________ Discuss mailing list Discuss@lists.snowdrift.coop https://lists.snowdrift.coop/mailman/listinfo/discuss
