How Many Active Users Are Really On Steemit?

in #steem7 years ago (edited)

I have seen a few random charts on active users currently on Steemit, but the data is hard to find and doesn't look right. It is subjective because each person may define an active user differently. And, as accounts age the active user percentage would change...I would guess it would slowly drop off. The purpose of this post is to show you the actual data for active users and what that drop off is.


What Is An Active User

This is just the definition I used, it's not something that is set in stone.

  • Account must have posted at least 5 times in the life of the account. This can be comments as well.
  • Account must have posted at least once in the past 2 months.

This part is subjective, but I think I was pretty lenient in this definition. The idea was just to try and eliminate accounts that never use the platform, bots people may use to upvote their stuff, etc.

Each column represents the month when the account was created.

  • July looks low because the month isn't even half way finished.
  • Looks like there was a lot of total account growth last summer, this summer looks even bigger!
  • The number of active accounts look like it's spiking like crazy.

This chart shows the total active users over time. This chart is showing that the number of active users has quadrupled in the past 3 months to a total of 38k as of July 13th, 2017.

Here is the same data, but this chart shows active users that signed up per month.

This chart is probably similar to what you may have seen floating around Steemit. I don't believe this chart, because my thinking was that as you progress in time, active users drop off. For example, a user that signed up on Steemit in June and is active today, may not be active 2 to 3 months from now. There has to be a natural drop off.

To find that out, I took snapshots of Steemit data every month for about the past year on the 13 of each month to see the active user trends.

  • The orange rows represent the month in which the account was created.
  • The blue columns represent the number of accounts that are active, inactive, and total for that particular month in which the account was created as of July 13th, 2017.
  • The green columns show the date at which I'm measuring account activity (1 post 2 months prior to that date). The percentage shows how many accounts are active from that particular account creation month.
    • For example: 29% (one of the cells highlighted in yellow) shows that 29% of the accounts created in June 2017 are active as of July 13th, 2017.
    • Another example: 3.2% (one of the cells highlighted in yellow) shows that 3.2% of accounts created in September 2016 are active as of July 13th, 2017.
    • Last example: 8.5% (one of the cells highlighted in yellow) shows that 8.5% of accounts created in September 2016 were active on October 13, 2016.

In the last 2 examples, the number of active users for the accounts created in September 2016 decreased from 8.5% active to 3.2% active today. This means some people stopped being active over those months...or maybe the accounts were setup for some other purpose besides posting.

It may be a little confusing, but you can see the percentages for each row decrease as you scroll to the left. Let's look at this graphically.

  • Each line represents active users from a particular signup month.
  • The Y axis is the number of active users for each signup month over time
  • The dates along the X axis is the snapshot at which I'm measuring active users.

Look at the light blue line (the line all the way at the top). This represents the number of active users that signed up on Steemit in July 2016. Over time, the number of active users that signed up in July 2016 has dropped from about 3,800 down to about 1,600. Each line is dropping off over time. Each line seems to be shrinking to about 1/3 of the original height over time.

This is the first chart I showed you, but hopefully it looks different to you now.

Each of the blue bars (active users) from recent months will shrink over time. The 18.5k active users that signed up in June may shrink down to 7k over time...same for July and May, etc.

This means that the 38k active user count as of today is inflated. What's the real number? Maybe it's closer to 20k if previous ratios hold up in the future. Time will tell, but let's assume for a second my 20k active users is correct. That means about 8% of total users are active. 8% is pretty good!

What is also really promising for Steemit is the number of total users signing up has grown significantly...there is no subjectivity in that!

If you read through all of that and didn't fall asleep, then you are as big a nerd as I am. Congrats!

All of this data was pulled using SteemSQL, please see posts from @arcange to read more about SteemSQL.

Sort:  

Great post.

I believe this pattern does not only occur in Steemit.

I believe that in all social networks the behavior is similar

For sure. Would be interesting to see this type of data from Facebook...assume they don't share it though.

That's crazy...not sure how I feel about that. Some regulations are good, but that's over the line.

@financialcritic Your criteria for active users are only for authors, content creators are much more rare than curators. You should use the criteria for authors and curators combined. Then you truly get all active users.

I did include curators that leave comments (at least 5 comments in life of account). If a curator just upvotes and doesn't interact then I am not so sure that it's a legit curator. Haven't really thought that out...so open to new ideas.

The whole goal of a curator is to use voting power, content creators don't need voting power.

I'm with you there, but still not following you. How would define a curator in a way I could identify in the database?

An active curator is someone who uses voting power at least X times in a certain period. The confusing part is that most accounts are hybrids that both post and curate.

Assuming that the curator isn't a bot, wouldn't that curator also leave an occasional comment on the post he/she upvotes?

I guess I'm just not imaging in my head a curator who is just upvoting without leaving a comment once a blue moon.

I like the idea of labeling as active only the users who invest a minimum of writting on steemit. On the other hand, many bots do leave comments on posts, usually asking for a follow or something like that. That said, maybe some bot's are counting as active and maybe some real active users are not.

That's true...randowhale/booster and all the welcome bots leave comments. So, maybe it's a wash?

It's a legit point, in my opinion. I know a lot of people who would never comment, but would upvote/like stuff. Ofc, there are fewer of those on steemit as you actually get rewarded for commenting, but they still exist.

I'll see if there is a way to filter those in somehow. And maybe it's a wash since there are bots who comment automatically.

Woah for the june 2017 sign up (active and inactive). It's interesting to see that last year, the booming months July, August, and September and then the dip. And it's looking promising for those months again this year. I wonder if there's a relationship between the months and the likelihood of more users flocking in to sign up at that time.

yeah, wonder why that is. Maybe people have more time in the summer? Maybe the average age is younger than I thought?

I'll keep my eyes peeled for in case you do a follow up for that. If they want to promote Steemit, those months could potentially hold even bigger schools of fish to pull into the community. Or it could be a coincidence but it's not looking likely as just that. Can't wait to see more of your take on this.

keep up this analysis. if the retention rate per cohort continues to decline faster than steem can add users, then this platform will run into the ring of fire problem. another important factor is the growth of engagement per user. i.e. are active users engaging with this product more by creating more content, commenting, reading, etc. This could mitigate the retention rate decline.

Great point...I hadn't even considered to look at rate of decline vs rate of new users. So far, the growth has been higher (based on chart 2), but it's something to watch...and maybe even compare to price of Steem.

I don't know what to compare this data to, but very interesting!

I think a user that at least post and comment in a number of x times in a month is active enough but a user that didn't post a thing in a month is not considered active for me.
The thing is that bots are active on steemit and actually affects data.

I agree with you, some feel differently. I guess technically I may one of the more inactive ones on Facebook...just using it to get info but rarely post. But I do post enough to be active even by my standards of 5 comments in life of account and 1 in past 2 months.

Statisticporn!

As a new user, trying to assess the future value of this platform...this info is really interesting. Really looking forward to seeing how this plays out.

I'd love to get my hands on this data monthly.

It's a great time to be on Steemit. I joined a month ago and thought I might be late to the party, but the party is just getting started based on these charts.

You keep bringing the good stuff! It's really interesting to see, I guess retention rates will go up in the future as the platform evolves. Now you kinda have to "look" for content, unlike other platforms who have mastered the art of "showing " you content.

This will probably also be solved with developers making their own platforms ontop of steem blockchain, each for it's own audience. For example it's natural that people who mostly care about photography share/consume in a different way than people who loves to go balls deep in statistics ;)

Yeah...wonder if retention will be higher after the last HF.

I have been struggling lately to find content also. Maybe I follow too many people and when I look to the hot/trending 80% is the same ol vacation rentals, youtubers cut/pasting their video, some one's vacation pics, etc.

Maybe I'll do some research to identify best authors to follow that write legit stuff and reply back to commenters.

Seems like there is a lot of curator projects going on at the moment, so I guess following those would be cool. Eg. a curator dedicated to certain topis. That also seems like a good business for the curator as you earn from your summarypost.

There is ofc a lot of good content here, just not everything is interesting to everyone. Seems like whatever content with the lowest common denominator wins 90% of the trending page :P

Coin Marketplace

STEEM 0.18
TRX 0.13
JST 0.029
BTC 57646.26
ETH 3029.18
USDT 1.00
SBD 2.26