You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more
Engagement is a toxic metric. Products which optimize for it become worse. People who optimize for it become less happy. It also seems to generate runaway feedback loops where most engagable people have a) worst individual experiences and then b) end up driving the product bus.
This reflection brought to you by the respected American journal which retweeted ragebait today... for the eighth time in the last two years. Did they do it because their social team knew that article would piss people off? Or does "sort by engagement desc" generate this magic?
I described the article as "clearly designed to be ragebait" in an early version of that tweet but in hindsight I think that might actually not be true. They could literally just have entirely conscientious journalists publishing daily and have their social team find the bait.
Maybe every editor wakes up every morning to serve the public interest. Maybe every journalist shines their light into every assigned story. Maybe every social media planner gives people what they want to see. Moloch cares not; he will subvert them to producing pollution anyhow.
So many product teams default to using engagement metrics to measure success, but there are a lot of products where engagement isn’t “success”. Think Headspace or Calm: isn’t the idea to use need those apps *less* over time, not more?
another example of this is the daily reputation cap we (well, I) introduced at Stack Overflow. This is telling you "yep, you've used SO enough today, time to work on other stuff as good programmers do." SO is not THE GOAL. Getting shit done IS.
Similarly @storyworth won't send more than one question a week. At some point that may change (for some specific use cases) but it's a healthy default.
"users" who talk the loudest and longest should never set your product direction
How do you weed that out? About to start a round of user interviews and could do with some input :)
if you select users randomly it shouldn't be an issue. If you let users self-select, you tend to get the most aggressive, talkative, opinionated users. These users may or may not have the right kind of feedback you need..
Isn't engagement everything on Instagram? Maybe people who optimizes for it are less happy, but it seems to me a solid product
A more basic set of metrics: consumer metric: enjoyment business metric: profit If you successfully maximize on both of those your result happy users and lots of profit. Engagement may be correlated with both enjoyment & profit, but not always.
Maybe your users are much happier with your product when they engage with for some max number of times/min per day, after which you find diminishing returns or even inverse-correlations.
An example might how some businesses have ruined user trust in their notification systems. For example I've never visited my linkedin and not had at least 2-3 notifications waiting; always spam. Users will pick up on this quickly...pic.twitter.com/81cUq9BnbS
So LinkedIn's strive for engagement was short-sighted; it worked at first (i'd check these notifications), but after a few times of them being just spam, it reduced my enjoyment (and ultimately my engagement went away too).
I agree with everything you said, but maybe LinkedIn got it all wrong. I see more people constantly checking their phone even if they haven't received a single notification. We are doomed by the dopamine hits of social networks and instant messages.
Yes I think they got it wrong. There is a rich literature from behavioral psychology on response entrainment. It turns out that when rewards are given periodically after some variable number of responses (VR) is when participants respond/engage maximally overall (ask any Casino).pic.twitter.com/DXZSq3GsoN
But provide the stimulus in the absence of reward (e.g. constantly notifying your users of spam message) will certainly result in response extinction. (i.e. when the conditioned stimulus is provided (a notification), but a reward doesn't follow, the dopamine squirt diminishes).pic.twitter.com/sNqCnLAjAW
When your product relies on dopamene infused attention. You are literally a digital drug dealer
I was disappointed in this "engagement hacking" in @stripe: no X to close the bubble -- you have to click a link to make it go away.
Someone made a choice to do that. Not a good look.pic.twitter.com/Kam9PZJGYG
Thanks for flagging—this wasn't intentionally missing the option to close. We just fixed it!
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
