DEV Community

Cover image for DEV: Followers != Readers
Maxim Saplin
Maxim Saplin

Posted on

DEV: Followers != Readers

Some dev.to analytics. There's a curios lack of correlation between followers and readers at dev.to. This week's stats is a good example:

Followers and Readers stats

Top comments (23)

Collapse
 
jonrandy profile image
Jon Randy πŸŽ–οΈ • Edited

DEV has a bad problem with bots. I get ~50 new followers almost daily, despite rarely posting anything - and most of them look like fake/auto-generated accounts.

@michaeltharrington ☝️

Collapse
 
maximsaplin profile image
Maxim Saplin • Edited

I also had this suspicion of many new followers being bots/fake accounts. Did a small research and it seems they might be actual users, the ones who just signed up. There's an onboarding screen for new joiners where they pick which topics they like and also who to follow. The problem is that not many of those who join dev.to will participate/return to the platform.

Btw, some of the new joiners/followers actually participated and commented.

Collapse
 
jonrandy profile image
Jon Randy πŸŽ–οΈ

Yup, all totally legit... πŸ€”

Junk followers

Thread Thread
 
maximsaplin profile image
Maxim Saplin

bangalorescorts is the best of the pack :)

Thread Thread
 
maximsaplin profile image
Maxim Saplin

Still, 6 out 12, not that bad...

Collapse
 
azasypkin profile image
Aleh Zasypkin

It's definitely not a bug; it's by design to improve the onboarding flow (which is why most new followers joined on the same day). See dev.to/ben/changelog-suggested-fol...

Collapse
 
maximsaplin profile image
Maxim Saplin • Edited

Still, there's a question of what kind of mechanics this follower feature brings, doesn't seem there's much use to it IMO

Thread Thread
 
azasypkin profile image
Aleh Zasypkin

Yeah, it's not immediately clear to me either. However, I'm very new to dev.to, so my stats look a bit different: half of the reader spikes correlate with spikes in new followers.

Collapse
 
mattiasfjellstrom profile image
Mattias FjellstrΓΆm

I've noticed this as well.

Go through new followers on a given day and you can see that most accounts are created that very same day. Many accounts include commercial for some random product in their bio.

Collapse
 
maximsaplin profile image
Maxim Saplin

At least the readers stats seem legit :)

Collapse
 
michaeltharrington profile image
Michael Tharrington

Thank you for the heads up! I've made the team aware of this one.

Collapse
 
kurealnum profile image
Oscar

Yup, same here. I've had a grand total of maybe 3-4 people like my posts, and I have 500 some followers. And as you said, a lot of them are very auto-generated accounts/bots.

Collapse
 
pcmagas profile image
Dimitrios Desyllas

Yeah that is an issue.

Collapse
 
dumebii profile image
Dumebi Okolo

I think why the followership doesn't correlate with the views is that there aren't a lot of active people on the website at any given time.

The method by which followers get notified of your articles is when they open their notification tab, unlike on platforms like medium where you get push notifications or emails notifying you of new posts by the writers you follow.
This is probably why a lot of the viewers come from Google, and not dev.to itself.

If dev can implement a feature like that, then it'd be cool.

Collapse
 
maximsaplin profile image
Maxim Saplin

Exactly the way see it, the followers mechanics is not complete and serves little purpose on dev.to

Collapse
 
dumebii profile image
Dumebi Okolo

Little, but not ent useless.
I think, with time, it'd get better.

Collapse
 
cicirello profile image
Vincent A. Cicirello • Edited

I'm going to give a different perspective on your graphs.

The graph of number of readers is showing the number of times in a day that your posts were read. That graph is increasing quickly to a peak of around 450 on 9/28.

I'd expect the number of reads in a day to correlate more strongly with your total followers than to your new daily followers. That second graph shows the latter. Your peak daily reads of 450 on 9/28 isn't just related to new followers who started following you on 9/28, but also those already following you.

I suspect that if you had a graph of total followers over time, it would have a similar shape as your readers graph. Although there are other factors that can affect readers as well such as if you made a new post that got extra exposure such as due to many reactions from other users and/or upvotes from mods.

Collapse
 
maximsaplin profile image
Maxim Saplin

Agree, that graph doesn't tell the full picture. Yet IMO it demonstrates the the issues with followers, the one observed not just be me..

Collapse
 
cicirello profile image
Vincent A. Cicirello • Edited

I agree. Definitely. Some who start following you when they sign up may never log in again. Or may ignore the notifications. Or they see the notification of a new post but don't click to read it. Or they then stop following you. Or some might be bots. If you have posts that rank well with Google you may get more reads from people that don't even have accounts than from followers. When you post might impact reads. And many other potential factors.

Collapse
 
chiragagg5k profile image
Chirag Aggarwal

Wow, interesting statistics.... I wonder why this is the case

Collapse
 
artxe2 profile image
Yeom suyun

I don't remember what it was like when I signed up, but it seems that new users are given random follow recommendations.
Since the views of a certain level or higher are mostly done by Google traffic through an unknown algorithm, there seems to be almost no correlation with followers.

Collapse
 
pengeszikra profile image
Peter Vivo

I think my blog post does not reach anybody, because my viewer numbers is really low. Any advice?

Collapse
 
maximsaplin profile image
Maxim Saplin • Edited
  1. Go to URL shortener and get a nice one for the blog post
  2. Advertise it at a Super Bowl :)