The Interledger Community 🌱

Discussion on: What's the Value of a Like?

Collapse
 
michaeltharrington profile image
Michael Tharrington • Edited

I feel like as a creator, I'd be super interested in "engagement" and particularly would like to know the metrics that define what "engagement" is. It'd be nice to see an overall "engagement" metric, but also to have the ability to get more granular and see what all the little algorithms are that factor into it. For instance, time on site seems like an important measurement for engagement. If your content is media like video or audio then time listening/watching via the media player would be one of the most important things to me. But there are many factors...

I think often engagement for articles on Forem (the platform this community is built on) would be measured using some combination of time on site, comment count, reaction count, etc. and an algorithm is used to define the heaviness (importance) of each. It's somewhat subjective to define the importance of different ways of engaging β€” i.e. should time on site be more important than reaction count and thus get a heavier multiplier in the algorithm? If you're presenting this info to a creator, I think it'd be ideal to give them a suggested algorithm for "engagement", but clearly define the parts of that algorithm and maybe even allow them to customize it so that they have the dials to define the different parts of "engagement" that are most important to them.

I hope this makes sense, haha!

Collapse
 
jon profile image
Jon Clement

I was recently reminded of the Klout score for influence - rip. Actually, their tech approach is a slightly interesting read: arxiv.org/abs/1510.08487
Agreed, in a lot of cases rolling everything up to one engagement value ie: 83 -- is useless. And yes, even if you include the details of the metric algorithm, these days, people require the ability to customize. I've been considering an adjustable "exchange rate table" to help gauge relative importances and the the conversation rolling.