@ozoned
impressions is the most accurate measure, most (if not all) platforms used it but then X changed it to "views" which is not accurate (your post are displayed but not viewed, unlike videos) and hence everyone switched to views

Engagement is a good metric but not as good as impressions, for example if you want traffic to your website impressions resulting in clicks are far more important than likes

@ozoned

https://mastodon.social/@wearenew_public/116556651501930786

@ozoned also we need to measure the value of the fact that checking in doesn’t devolve into endless scrolling - I like that it only takes a few minutes of my day to connect, see a few new things and be on my way

@ozoned

human boosts are the most important metric. an actual person is willing to stake their following on suggesting something you created is worth other people's time. that's huge and what the platforms are moving away from with algos. they know it's a big ask and they've hitched their wagon to virality but fake it with robots. we are the mirror to their BS.

culture was built on word of mouth. a trusted web of recommendations. nobody paid off, no way to buy attention. we threaten them.

@ozoned One thing that's nice about the Fediverse that's not quantifiable, is that I've actually made some friends here that I've never met in the real world.
In all the years I actually was on FB and looked at the "feed", it was zero.
(There have been a couple--ok maybe just one--from FB groups, but in the feed? nobody).
In the days of email lists, I had friends through those. That was pretty much destroyed by social media, mainly FB (and this is why I will curse meta forever).
Anyway, like I said, this is important, but not directly quantifiable.

@ozoned Thinking this over. If I want to measure what I'm getting out of the Fediverse, one way to look at is, if I post something, does anyone care? And do I care about what other people are posting?
A good measure of this is how many people actually comment on my posts, and how many posts do I comment on.
You could quantify this as "median written responses per post". Note I'm saying written responses, not boosts. Boosts are not enough effort to count.
The reason I'm saying *median* is that some networks have accounts with huge fanbases, and those followers do respond. That's "engagement" useful for marketers to know about influencers. And we do have a few very-well-followed accounts here. Using median means these outliers won't skew the result that much.

I mean, if I post something, and two people say something back, I'm happy.

@ozoned for all that my last couple of years of education have involved some training-wheel level Data Science and Research Methods class work, I’m not sure this is a thing anyone has worked out. I feel like it COULD be done and I can imagine the shape of it, but it would involve rethinking what we consider a metric at a very fundamental level.

With social “media”, the subject of all the metrics we have is NOT the person … it’s the product, the individual units of “content”.