An active user should be defined by a set of conditions.

For this case study, I considered the following

1. Transactions per week
2. Transaction volume per week
3. Programs interacted with, in that week

1 point each was awarded for

• transactions/week > 5
• transaction volume / week > $10
• interacting with 5% of the total non-system programs on eclipse; bringing the highest possible score to 3 - with 0 indicating an inactive user and 3, an active one, as well as, barely active users and moderately active ones

For the Eclipse chain, I noted that the 'barely active' group is mostly occupied by those interacting with turbo tap

Of course, these heuritics might vary across chains depending on the purpose of the chain and what is seemingly more important to them. (Other considerations not covered here; weekly stakes, weekly liquidity provision)