1. Approach 1.0
   1. 1. Influencer sends DM to its top 1000 followers - with link to register on socialexporter.com
      1. 2. These followers then participate in the leaderboard game and send DMs to the influencer's followers to increase their score and win the bounty
         1. Twitter doesn't allow you to send dms to people who you don't follow unless they have opted to do so.
            1. After testing this approach on 1000 followers, I found out that only 90 people had opted to receive DMs from people who didn't follow them
            2. Problem - the follower sending DMs other followers didn't seem to be working
2. Approach 2.0
   1. • While the follower can't send DMs to another follower, it can tag that follower in his tweet
      1. For example
         1. Day 1 - Influencer sends DMs to 1000 followers - Out of those 1000 followers - 100 participate in the leaderboard game
            1. Day 2 • every follower posts 200 tweets tagging 3 followers in the tweet with his referral link • 200*3*100= 60k other followers reached
            2. Day 3 • 6000 more followers participate • post 200 tweets each with their referral links - tagging 3 other followers • 200*3*6000 = 3.6 million followers reached
            3. • Twitter allows every user to post 2400 tweets in one day • It depends on the competitive spirit of the participant as to how many tweets he wants to send • he can customize that on the socialexporter.com/soporter/sendTweets
            4. • However, there's an app limit of 300 tweets per app per day • This can be removed by getting the 'elevated post status' as explained in the Findings map
3. Conversion Rate
   1. • Conversion rate is the measure of followers who actually signup and participate in the leaderboard bounty once they have been notified of the socialexporter.com
      1. • The conversion rate of an influencer sending the DM - A • The conversion rate of a follower marketing to another follower - B
         1. • Influencer sends DM to x followers - Ax are converted • These Ax followers tweet about it and tag another y followers - Ax*B*y are converted • These ABxy further tag another y followers - Ax*By*B*y are converted - AxB ²y ²
            1. • The exponential increase is evident with this approach
            2. • Therefore it becomes irrelevant to rank all the followers according to some critieria and then send DMs/tweet • If we can find the top 1000 (or 5000) followers who engage best with the influencer, then we don't have to parse through the entire follower list
            3. • Hence, comes the ranking algorithm to get best engaging followers without parsing through the entire follow list of thousands/millions of followers
4. Ranking Algorithm
   1. To get the most engaging followers, I used the following parameters - 1. all the followers who have retweeted the posts of influencers 2. who have replied to his tweets 3. who have been mentioned by the influencer - mutual engagement 4. have excluded verified accounts as they're least likely to participate in the bounty game
      1. • For now, i have assigned equal weights to the first three parameters • After, running a few dry runs and looking at the quality of the top followers, different weights can be assigned