Social Farming - Discussion

Ideas
#1

One of our goals in the roadmap is to incentivize our community to engage on telegram, twitter and on our forum by giving $SNAFU as a reward.

For now, this is our plan:

  • Use SourceCred for our forum and give a cashback of $SNAFU based on the user reputation
  • Use Combot for Telegram and remunerate each month the more active users
  • For Twitter, instead, we want to select for each post a winner from the replies, one from the retweets, and one from the likes and give that user a $SNAFU reward. This would happen after 6 days of each post.

Telegram

With Telegram we have this type of information from Combot:

Name ACT MSG AD DSJ Joined Last message
Dan 49.35% 124 7 8 2021-05-29 2021-06-06
Eth Man 24.89% 36 3 2 2021-05-29 2021-05-31
Reem 9.17% 55 9 9 2021-05-29 2021-06-07
777fanno21 6.92% 27 3 8 2021-05-29 2021-06-06
David (Please DYOR…) 4.03% 19 2 1 2021-06-05 2021-06-06
Masnat 1.75% 3 1 0 2021-05-30 2021-05-30

ACT is based on the length of the message
MSG is the # of messages
AD is active days
DSJ is the number of days after entering the group

We can filter this list from/to date, so we need to come up with a formula to use for distributing $SNAFU.

Also, should admins be remunerated?

Twitter

Our idea was to give 10 $SNAFU per winner, so 30 $SNAFU per post, but let us know what you think about this.

Forum

There is already an ongoing discussion about the Source Cred integration here: SNAFU SourceCred Instance

As per the roadmap, Social Farming was set to be released in July, but if we are able to close this conversation earlier we can deliver it earlier.

3 Likes
SNAFU Roadmap 2021
#2

yes admin should be remunerated.
and I also agree that more active users should receive an incentive.

1 Like
#3

HI. i believe that Twitter can be very useful to grow our community… so any rewards for spread the word and Retweet will helpful. If ta budget is 30 $SNAFU per post, i would give 1 to first 10 and 20 to retweet with most likes. ( that’s exclude any draw/lottery)

#4

You could use SourceCred to incentivize chat as well through Discord. I am personally not a huge fan of Telegram, but I know it’s very common in the crypto space. Discord is just more flexible and you can create channels made for specific types of conversation. In Telegram I find things get lost. Discord helps to organize discussion especially as the community grows. Right now it might be easy to follow the conversation, but imagine when there are thousands of people in there. It just gets messy.

I think Mods should be remunerated. I would be in favour of funding a mod budget from the treasury as well.

1 Like
#5

Hi.
I shear the same opinion about Discord/Telegram. There is a plan to move to Discord when community will grow a bit bigger.

#6

I feel like it is always easier to move when the community is small, and it just sets a standard for new people that join the community. Just my 2 sats

1 Like
#7

I think mods should be rewarded for sure. They go a long way in building community and will probably put in much more work than the rewards equal.

I can’t comment on COMBOT and Telegram versus Discord and Sourcecred as I just don’t have the experience to qualify me. I have found that each platform has distictly different audiences however. I belonged to a project with both and we had people who stuck to either one, and disliked the other. Generally Discord attracted more project & detail-oriented folks and Telegram attracted folks focused on more general messages who didn’t want to go deeper into things. They were surprisingly different crowds.

With regard to the twitter rewards I would like to emphasize that writing a thoughtful reply is 100x more work than clicking a ‘like’ button or a retweet button. With that in mind I suggest you reward the most liked reply with 15, and the other two with 7.5 each. Or maybe even 20/5/5 but that feels extreme.

One last comment is no matter how you format it people will try to game the system. That is, they will try to exploit the social farming rewards in every way possible. Sad but true.

Cheers!

1 Like
#8

Hi. I agree with you that best performing comment/retweet should have highest reward.

#9

Ok, so we are having problems with Twitter ads based on their crypto policies. Maybe we’ll have same problems with Instagram.

So we need to accelerate at least with the Twitter farming. I agree that comments should have more reward than likes and retweets. I will put down some numbers and post them here.

1 Like
#10

Allocation per tweet: 300 $SNAFU

Super Prize

  • One random person from comments wins 50 $SAFU

Prizes

  • comments have 50% of $SNAFU
  • retweets have 25% of $SNAFU
  • likes have 25% of $SNAFU

Example

Let’s say we do a tweet and we have 25 comments, 40 retweets, and 50 likes.
5 days after the tweet we choose a random comment and give 50 $SNAFU.

We now have 125 $SNAFU for comments, 62.5 $SNAFU for retweets, and 62.5 $SNAFU for likes.

CommentRevenew = BUDGET / # COMMENTS = 125/25 = 5 $SNAFU
RetweetRevenew = BUDGET / # RETWEETS = 62.5/40 = 1.56 $SNAFU
LikeRevenew = BUDGET / # LIKES = 62.5/50 = 1.25 $SNAFU

If I commented, retweeted and liked then I will receive around 8 $SNAFU.

What do you say?

#11

Hi.
The system looks good the only question is what will be total allocation on $Snafu tokens to this promotion and for how long its going to run. I would change the reward of 50 $Snafu, instead of random reward to best performing. So the comment/retweet which have most like/retweet/comments or use engagement score.

#12

What worries me is how to track all of this. For example for random picks there are a lot of instruments online so it’s easy. But here not only we’ll need to have a list of all appliers and see if they retweeted liked (same as bounty basically) but we’ll also then need to check the retweets and likes on their retweets as well.

If we could somehow simplify this it would be better. I like the idea of randomly choose people because it’s more easy.

For the budget: yeah we need to allocated a budget and to understand how many tweets it can handle. I think answers to our own tweets should not count in this.