STEEM Churn Report Q1 2019

in #utopian-iolast year (edited)

Churn and retention are facts of business. You will win new users and you will lose users. Churn measures how many users you are losing. Whereas retention measures how many users you are keeping.  Churn can be calculated by the following formula

Churn = Number of Active users at the beginning of the period + New registrations – Number of Active users at the end of the period.

Churn rate can be calculated using 

Churn rate = Churn / Number of Active users at the beginning of the period.

A high churn rate on a social media/networking site would indicate low engagement lack of loyalty and high dissatisfaction.   To give us some sort of base to use as a benchmark for STEEM, I have taken this chart from vertoanalytics.com

Repository

https://github.com/steemit/condenser  

Aim of Analysis

The aim of this analysis is to calculate the churn rate on STEEM as a social networking type blockchain* for Q1 of 2019 and compare this to the previous 12 months

*STEEM is not only used a as a social networking type blockchain.  Some people are investors and wallet holders only.  Therefore, I have taken figures for all new registered accounts and I have also taken figures for those that have either poster or voted or both and excluded those that have not.

I also wanted to look at different account creation routes to see if there is any difference in retention from route to route.

All data for this analysis was taken from Steemsql held and managed by @arcange.  As to not distract from the data, I will first present the findings and then the queries used.  

Monthly Churn

    

The table and chart above show the monthly churn and churn rates including ALL registered accounts as new accounts, no matter what their actively level on the block.

It is clear to see from this that the churn increased in January to reduce again in February and March.  This increase may have been down to the backlog of accounts from December that were set up in January.

The average monthly churn rate for 2018 was 39% and the median is 35%. The average monthly churn for the first 3 months of 2019 was 24% and median 25%

Not all accounts became active in a social sense.  Some accounts are just wallet holders.  To reflect this the table and chart below have an adjusted in the new account’s column.  This column now only includes accounts that have either voted or posted.  Churn and churn rates are calculated the same as above.

   

From this we can see the reduction in churn rate seen during 2018 continued into the first 3 months of 2019.  A churn rate of only 6% was recorded for March.

The number of distinct users has been increasing since December with March seeing the highest number of distinct users in the last 4 months.  This coupled with the lower new (socially active) accounts resulted in a very low churn rate.

courtesy of @arcange

All in all, monthly churn rates when you only include accounts that either/or post/vote show a lower churn than that including all accounts.

The average monthly churn rate calculated this way was 9% and the median was 10%.

Quarterly Churn

Month on month churn is not the most effective measurement as a user might not be active one month but become active again the next month.  Therefore a better way to calculate churn is on a quarterly basis.

  

The table and chart above show quarterly churn where New accounts includes ALL new registered accounts. Q3 & Q4 of 2018 showed churn levelling off. However, Q1 of 2019 shows a slight increase in the churn rates from 52% to 53%.  

The table and chart below show quarterly churn for socially active accounts.  Although we seen an increase of 60% in the number of new socially active accounts in Q1 2019 over Q4 in 2018, we also see a further drop in the number of distinct voters and authors.  However, as the drop in distinct users was not as severe as the increase in new active accounts, we see an overall drop in churn for Q1 to only 29%.

  

Retention by Creator

HF20 introduced methods to create accounts with account creation tokens, or by burning Steem, Opening up the doors to alternatives to Steemit Incs account creation service.

With this in mind I took a dive into drop offs to see who was retaining the most users they created.

Since the 1 October 18 Steem has 104K new accounts.  Of this almost 16K, or 15.2% were active in the last 14 days.

Looking at social activity, as in those that posted or voted, 41.8K new accounts became active between 1 Oct and 31 March.  Of this 37.76% were active in the last 14 days.  This 37.76% represent retention of accounts set up during this period only.  This would suggest a churn on new accounts since October of 62.24%

The average drop off day for accounts that were not active in the last 14 days is 27.49 and the media is 14.

  

The line chart represents the drop off day for accounts that were set up during this 6-month period and the table beside that shows details per account creator.  Steemit Inc being the largest account creator retained 35% of all accounts that became active, with the median drop off day for the remaining 65% being 14.  @steemmonsters retaining 67%, @blocktrades 34%,  @steemhunt 41%, @oracle-d 47% and @partiko 42%. @comedyopenmic also looks good in there with over 46% of new accounts becoming socially active and retaining 54% of them.

Conclusion

Looking at STEEM as a social type platform, it would be my opinion that only including new accounts that became active on the social side is a fair representation of churn on steem as a social platform.  Some accounts are wallet holders (accounts with active transfers in or out of the wallet and no social actions), some accounts never had any activity at all.  Therefore, there is justification for excluding these accounts from the calculations when looking at churn from a social aspect.

However, it is worth pointing out, accounts that have ONLY JSON transactions (such as steemmonster players) would be included in accounts that never posted or voted.

Although overall churn rates are dropping, the churn calculations factor in accounts from all time.  Therefore I feel the churn calculated on new users since October only, based on 6 months with activity in the last 2 weeks of 62% to be a more reflective position of churn on steem.

It's interesting to see the retention by creator and it's awesome to see the %s for @comedyopenmic

The Data and the queries

As mentioned I used Steemsql and PowerBI to gather and model the data.  The query used to get the unique voters was

SELECT voter, timestamp FROM Txvotes (NOLOCK) where timestamp >= CONVERT(DATE,'2019-01-01')             and         timestamp< CONVERT(DATE,'2019-04-01')

With the data returned from this query, a distinct count of voter (for both month and quarter) was calculated.

The query used to get the unique authors was

SELECT author, timestamp FROM Txcomments (NOLOCK) where timestamp >= CONVERT(DATE,'2019-01-01')             and         timestamp< CONVERT(DATE,'2019-04-01')


With the data returned from this query, a distinct count of author (for both month and quarter) was calculated.

To get the total unique users, I combined both of the above queries and then ran distinct counts on the name.

The query used to get the number of new accounts was

Select name, created FROM Accounts (NOLOCK)

However, this query only gives ALL new accounts.  To get the distinct counts of new accounts that posted and or voted I used data from a different report.  This is the monthly new user report that is published on steem under my account.

The query used to get the accounts creator was

Select * FROM txAccountcreates (NOLOCK)


Sort:  

Hi @paulag, great update on the churn rates! Seeing yours and aracange's graphs I realized that we seem to have reached a pretty stable number of monthly active users. I think this is also quite positive news after the steady decrease over most of the last year. The "number of accounts by drop-off date" graph is an interesting one, great addition to the churn rate graphs! I'm surprised to see a clear peak at around day 40 - do you have any idea what happened there? Steemmonsters retaining 67% of their accounts is quite good, also compared to the other account creators.

Your contribution has been evaluated according to Utopian policies and guidelines, as well as a predefined set of questions pertaining to the category.

To view those questions and the relevant answers related to your post, click here.


Need help? Chat with us on Discord.

[utopian-moderator]

Thank you for your review, @crokkon! Keep up the good work!

I am not shocked to see the retention levels dropping. Steem has been so low that users joining can't be chasing riches compared to the users joining at the all time high last year. I would like to see more ways of trying to get users to stay by explaining this place in lay man's terms so it is simple and basic. The platform is still way too complicated for the average person in the street.

agreed in full, and as I mentioned in a comment below, when people find out how much there is to learn, the actual run away faster.

I wonder how much some of the churn could be mitigated if those coming were forced to read how the platform works, and maybe answer a questionnaire showing they understand that the interfaces such as Steemit are not the chain itself. I think some might stick around if they were to understand as time goes on there will be many ways to access the chain and benefit from what it offers.

Lots of misinformation out there (especially on Youtube), FUD from crypto purists who don't understand this variation (usually because to them crypto is meant for pump and dump). I was one of the lucky ones to encounter some of the legends here who break things down. I suspect many are not so lucky.

Nice job on the numbers and picking the criteria to measure by.

so I have some interesting insights. I give some free delegation to people that take a course on about steem. It looks to me when new people see how much there is to learn, the run faster than when they didnt know there was so much to learn. In fact, I am very much considering changing the target student from newbie steemains to enterprise type users.

Those retention rates from those communities are very encouraging and demonstrates that the underlying flexibility done for HF 20 have made an important difference. I think that the missing link to improve even more is a front end with Hivemind being used to its full potential!

Posted using Partiko iOS

I think both retention and acquiring new users are a problem. with any luck, the changes from HF20 just need a little longer in play to see improvements, especially from app owners

Really enjoyed this post I know there are very few of us who like to look at the user numbers but I happen to be one of them! A little concerning that we average say new 2500 active users per month!

We really do have an retention problem and I think it’s due to the over reliance on the earning money side of the platform and not the just use it for all its other benefits and not just money

I think now that steemit.com is positioning itself to allow community commits we may be able to give the bulk of new users a better experience

Also surprised to partiko not on the list

Posted using Partiko iOS

"We really do have an retention problem and I think it’s due to the over reliance on the earning money side of the platform and not the just use it for all its other benefits and not just money"

I could not agree more and I have been thinking about this a lot the last while. In my head, people need to start seeing steem as a value-added benefit to existing models and businesses. To me steem is an awesome platform for enterprises. Enterprise being anything from blogging to gaming, but with any enterprise, its so much easier to get started if you have your own start up capital. But that's a whole different post :-)

I only just found this post know, after asking Asher if there was a way to find out how many newcomers we are losing. From what I've seen these last couple of weeks, I assumed that number was huge.

I'm no good with stats - my head is spinning, after trying to read the graphs and stats above. Stats always confuse me, lol. I really need to learn to read them decently.

I will not make conclusions from stats without asking for confirmation to see if I understood correctly.

So, did I get it right that these last 6 months, we lost about 62% of the newcomers, on average? In the last 3 months, only 29%?

I'm probably not reading it right, because that last number is like the opposite from what I had in mind...

This post has been included in the latest edition of The Steem News in 10 posts - a compilation of the key news stories on the Steem blockchain.

I am with the comments below, The pat form is a bit complex for the average person to pick up and be like, OH, ok. 3 different money formats, savings, then you can only vote and post so many times before you have to recharge, go tot learn Mark up too...

Steemit is to geared for the tec market and not so friendly for the masses .

Till we have a better posting method and an interface upgrade (something that lets you see a visual layout for how many posts / comments you have left) and some other tweaks (like blocking notifications from drug war battles)... I think we wont see the BIG push in growth from the user base.

All that said, I like the platform so I am staying. :)

A mere 6% last month, that's rather fantastic I'd say.

Just a pity about this
" This would suggest a churn on new accounts since October of 62.24%"

That period covers the worst of the bear market though, which seems to be at an end (touch wood).

wood is touched, lets hope the only way is up :-)

Hi @paulag!

Your post was upvoted by @steem-ua, new Steem dApp, using UserAuthority for algorithmic post curation!
Your UA account score is currently 7.031 which ranks you at #89 across all Steem accounts.
Your rank has not changed in the last three days.

In our last Algorithmic Curation Round, consisting of 211 contributions, your post is ranked at #15.

Evaluation of your UA score:
  • Your follower network is great!
  • The readers appreciate your great work!
  • Good user engagement!

Feel free to join our @steem-ua Discord server

Thank you so much for participating in the Partiko Delegation Plan Round 1! We really appreciate your support! As part of the delegation benefits, we just gave you a 3.00% upvote! Together, let’s change the world!

Great data @paulag :)
Really liked the part retention by creator. @steemmonsters is doing a great job again.

yep @steemmonsters is doing an amazing job

Hey, @paulag!

Thanks for contributing on Utopian.
We’re already looking forward to your next contribution!

Get higher incentives and support Utopian.io!
Simply set @utopian.pay as a 5% (or higher) payout beneficiary on your contribution post (via SteemPlus or Steeditor).

Want to chat? Join us on Discord https://discord.gg/h52nFrV.

Vote for Utopian Witness!

hope that the figures turn the other way around soon, would be nice to see some growth. The new account creations are hopeful though