Enrollment Fees & Kernel Funding
-
what are initial reactions to:
Requiring $50-75 enrollment fee when accepted to mempool
Requiring ~$500-700 enrollment fee when accepted to fellows track?Still with our extensive scholarship program.
-
@aliyajypsy said in Mempool:
Requiring $50-75 enrollment fee when accepted to mempool
Requiring ~$500-700 enrollment fee when accepted to fellows track?This shape is a big step towards making Kernel education free, so that’s my first thought. I want to defend the open access we’ve created.
I like the idea that there’s some small financial commitment to apply. Even a $3 charge for a meetup makes a big difference - the turnout much higher and people more focused.
That takes me to refundability. Making it refundable is an incentive for people to opt-out of the mempool, which is helpful for us too. It keeps the mempool cleaner and gives us a focal point to learn why people disengage.
Putting these together, my instinct is to make the mempool fee higher, say $150, but anybody can self-qualify for financial assistance and get in for free. We can also use all these mempool fees for a scholarship pool for fellows, so we don’t need to depend on outside scholarships as much, and increase the open access to fellows.
Next thought: 15 fellows per week X $500 = infinite runway! (~$30k per month) Kinda magical how that worked out. @cryptowanderer did you think of that when we were designing the mempool shape and looking at rates?
-
my initial inclination is to say that to receive a refund requires a max 10 minute conversation about why you’re dropping out / what feedback do you have for the program? a steward can conduct the brief interview.
but also, can’t coerce. maybe leave it optional? just trying to ensure we get the feedback to improve and better serve
-
@aliyajypsy From my experience, I usually get refund interviews after the fact.
The pattern comes from early-stage product launches. At first, you’re more interested in watching customers use your product to see how they use it, if it’s useful and why. You charge money - not for the revenue, but to filter out the segments that see value up front and engage (not the tire kickers). But you also want to see strong signals of why they defect and want to make it easy, so you can hone in on issues that might be significant but not deal-breakers. In that context, a refund-on-request policy helps. Usually, it’s just an email back saying, ‘Your refund has been sent. Please let me know you got it. Also, just wondering if you might share why this didn’t work for you? It’s helpful for us to improve.’ They’re usually more open and conversational about why than if it’s a requirement up-front. Then they just give you some one-liner.
-
-
Thanks for starting this and all the ongoing work @aliyajypsy - it’s really awesome. @saintsal
- A fee to apply may make it prohibitive for people to continuously re-apply, which I don’t think is great.
- Are we all talking about an enrollment fee (i.e. paid once accepted to the mempool track)? If so, it could be helpful, especially if refundable. I like that it would go into a “scholarship pool”.
- I don’t think it’ll give us infinite runway (one of the risks being we will need to take breaks occasionally) but $500 per fellow should help significantly.
Now, how do we implement this? Of course, my mind goes to FreeLearn. Though it doesn’t quite fit, for various reasons. It would be simpler to:
- Write a mempool contract that can receive DAI (lock it in a Yearn vault for slightly more return + much more contractual complexity).
- Allow anyone who sends DAI to withdraw what they send at any point.
- Allow the Kernel multisig scholarship fund (we need to set up another multisig anyway, for better accounting, greater security etc) to sweep funds in the mempool at periodic intervas.
The third point is scary and not good design. However, if we communicate clearly once the mempool is closed that we will be sweeping funds in it to go towards fellow scholarships, and then do the sweep ~2 weeks later, just before we open apps again, that could be OK.
One advantage of this is that it also gives us a (partial) source of truth for an ACL. It is partial due to self-qualifying financial assistance. We could include another function for that, which just requires people to register.
However, (i) this still costs gas and (ii) if we use the mempool contract as an absolute SoT, then anyone can “register” or send 150 DAI and we would mistake them as being in the mempool track.
