February '19 New STEEM Account Report

in #utopian-io5 years ago

Each month I will be producing a report with details of New accounts created on STEEM.  It has long been established that the account creation route via Steemit Inc has flaws, with delays, batch account creation and users never activating accounts once they have been created.

At one stage Steemit Inc acknowledged this problem and issued a post stating they have put in place measures to track every stage of the sign up process, from first landing on the sign up page to logging into their newly created accounts.  However, we have seen no further information from Steemit inc to see where the exact problem is.

HF20 implemented changes to account creation.  These changes made is easier for others to create accounts using free account creation tokens.  However, this process still requires delegation, so accounts have enough RCs to transact.  There was also a change in accounts created with fees where by under the new system, the fee would be burnt.

If you missed last months post you can check it out here

https://steemit.com/utopian-io/@paulag/january-19-new-steem-account-report-1549018740942

Aim of Analysis

The aim of this analysis is:

• Establish how many new users registered

• Establish what % of accounts have posts and voted and are just wallet holders.

• See who is claiming account creation tokens

• Establish if delegation is been given to accounts so they can transact. 

 Repository

https://github.com/steemit/steem

Overview

Feb 18

 

January 19

 

February 19

   

Feb '18 was a great month for account creation with over 72k new accounts created.  The crypto crash of 2018 had a major impact, but we are now looking at a decline in the drop of new accounts and things seem to have levelled off.

Feb19 shows Steemit inc again creating the bulk of new accounts (85.31%).  Although the total accounts created is down on Jan, this is of no consequence as Jan included a backlog from December.

47% of accounts created in February have never transacted on the blockchain via wallets or posts or votes, this can be compared 58% in Jan 19. A guess on the reason for this reduction is that many of the backlog accounts from December, set up in January by Steemit inc, due to the delay, never activated their accounts in the first place.

Another notable different between Feb 18 and Feb 19 is the Creation fees and creation delegations.  This change is down to HF 20.

Two interesting values I keep an eye on are the % about profile complete and the % of accounts with website on their profile.  These values give an indication of the calibre of user.  Those with intent of using the platform will have more tendency to complete their profile.

@oracle-d had a fantastic month for account creations which I would put down to the implementation of @steem-ninja, in Jan they created 0.76% of new accounts and this rose to 2.55% in February.

For clarification on the Account activity levels

• Fully active accounts have both posted and voted

• Not active accounts have not vote/posted or made a transfer

• Wallet users have not vote or posted but have made a transfer out from their account.

Activity Level by creator

Taking the top 6 account creators, below we can compare the activity levels of the accounts created.  @Partiko stands out with the highest % of accounts not active in any form.  @Blocktrades and @Oracle-d stand out for the highest % of wallet users and Steemit Inc stand out with the most accounts that have voted only. 

  

HF20 Changes & Delegations

Under HF20 accounts can now be set up by either burning the STEEM and sending delegation so the accounts can transact, or they can be set up with discount account creation tokens where the fee is 0, however these accounts also need delegation to transact.

  

97.85% of new accounts set up, were set up with the newly introduced account creation tokens.  The balance, 413 accounts, were created by burning steem at 3 steem per account.

However, when accounts are set up, without delegation, there resource credits and voting mana need to be boosted so they can really transact on the block.  So many new accounts have received delegation?

15,286 which is 80% of accounts set up in January have received delegation.

Of this 15K account that received delegation, just over 96% received 30K Vests which we can see in the pie chart below

The first table below shows the new accounts, who set them up and how much in delegation they received.  The second table shows the people setting up accounts and how much in delegation they account they set up received.

   

Account Creation Tokens (Year to date)

The introduction of account creation tokens in HF20 allows users claim the right to set up an account.  In total 61K account creation tokens were claimed in January.  Below shows who has been claiming these account creation tokens.  We can see Steemit Inc has claimed 19% of these, yet they have set up 87% of new accounts year to date.

       

Conclusion

2018 seen a down trend each month on the number of new accounts created.  Due to the standstill of accounts created by Steemit Inc in Dec and the catch up in January, its hard to tell if this trend has bottomed for sure, but it is looking positive.

Although there are many new account creation services, such as Steem-ninja, we have yet to see a large shift in the % of accounts set up by Steemit Inc v’s others. Delegations are still required so new accounts can transact. Steemit Inc has the fund to do this so it is still favourable for steemit inc to be the main account creator.

During this analysis I also noted a change in account creation habits for steemit inc.  In 2018 Steemit Inc would create a bulk of accounts once a week.  February seemed to be a lot smoother on a daily account creation rate. 

The Data-source and Query

As always to produce this report I connect to Steemsql (paid subscription services held and managed by @arcange ) with Power BI. The SQL query used to gather the data for this report was

Select *
FROM Accounts (NOLOCK)
where   
( created >= CONVERT(datetime,'01/01/2019')

To get details of who set up each account I used the following query

Select *
FROM txAccountcreates (NOLOCK)
where   
timestamp >= CONVERT(datetime,'01/01/2019')

and to get details of accounts that used their wallets I used the following

select [from], timestamp
from TXTransfers
where   
timestamp >= CONVERT(datetime,'01/01/2019')

The query used to find out who is claiming accounts was

Select *
FROM TxAccountClaims (nolock)

Once I loaded the data into Power BI, I then carried out transformation and calculations using DAX language.  

 Shameless Promotion

 Like what we are doing?  Dont forget to vote for @steemcommunity as your steem witness.  You can vote for our witness using SteemConnect here: https://steemconnect.com/sign/account-witness-vote?witness=steemcommunity&approve=1  

Sort:  

Hi @paulag, thanks for the update on the account creation process!
Around 19k new accounts isn't too bad. Without knowing how many accounts were requested, it's hard to say if the Steemit creation process is a limit here. I agree that January likely had some backlog from December. The changes for blocktrades and oracle-d is quite interesting, blocktrades almost halved their number while oracle-d more than doubled it. It is true that partiko stands out with the highest number of inactive accounts, but seeing this metric coming down from ~85% last month to ~61% is again a good sign, IMO.

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 @crokkon for the awesome review, it was good to get back to a bit of analysis. Oracle-d had a bit of a push with the new referal service and steem-ninja. Partiko are story adult accounts so that could be the reason for that. Im looking forward to seeing march figures to see if we have infact now bottomed in terms of numbers of new accounts

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

Really great to read up on this stuff @paulag! Thank you for taking time to put it together! A real fact based look at account creations! Need more of this type of work on the chain. Stay tuned for @steem.ninja for more coming, including live event account creation, fiat to Steem credit card payments and voucher purchase system!!

@steem.ninja is rocking it, nice work. great to see new accounts being set up from others than steemit inc. it will be interesting to look at retention of accounts from different creation routes

Wow, what an excellent collection of data, thank you :)

I thought Steemit has only small population.Now I know, there are lots of people around the globe and the new members are so welcome to this blockchain.
God bless Steemians. Thank you Ma'am @paulag for the collection of these data.

Posted using Partiko Android

small but powerfull, we are a global community full of passion, thanks for stoping by @olivia08

You are so very welcome.
I love to see how you did a great job here.

Posted using Partiko Android

This post has been included in the latest edition of SoS Daily News - a digest of all you need to know about the State of Steem.

Congratulations @paulag! You have completed the following achievement on the Steem blockchain and have been rewarded with new badge(s) :

You got more than 6750 replies. Your next target is to reach 7000 replies.

Click here to view your Board
If you no longer want to receive notifications, reply to this comment with the word STOP

Do not miss the last post from @steemitboard:

Carnival Challenge - Collect badge and win 5 STEEM
Vote for @Steemitboard as a witness and get one more award and increased upvotes!

Very good research and information, thank you for this

Posted using Partiko Android

you are most welcome @crypt-skip

Interesting data. It seems that only 10-20% of the new accounts are used fully. I wonder why.

its a problem that is impossible to get to the bottom of right now. much of my specualtion is the time wait from signing up to getting your key, people lose interest. then there are people just storing account names too, there seems to be a number of factors

Hi @paulag!

Your post was upvoted by @steem-ua, new Steem dApp, using UserAuthority for algorithmic post curation!
Your post is eligible for our upvote, thanks to our collaboration with @utopian-io!
Feel free to join our @steem-ua Discord server

Taking the top 6 account creators, below we can compare the activity levels of the accounts created. @Partiko stands out with the highest % of accounts not active in any form

My guess would be that Partiko gets a lot of traffic, in part, from people who already have steem blockchain accounts but want a way to read remotely. They may, in fact, being far more interested in just consuming content than configuring a profile or voting, just based on when and where they use a mobile device as opposed to the main platform.

15,286 which is 80% of accounts set up in January have received delegation.

In a sense, this is the best news in the whole report. It means that at least 15,000 accounts were created with the knowledge and intent that they do something on the blockchain. That definitely describe some intentionality, no matter what it is.

The March report is going to be make or break for understanding what the creation trends are.

All accounts set up by steemit inc get delegation (15sp or something) and as 85% of accounts are set up via steem inc, its no sure way to know they have any intent. Infact steemit inc produce the most amount of voting only accounts, so the intentions of vote farming from free delegations could be the only intent.

You know it's a terrible world when I actually look for something positive to say and you shoot me down. Legitimately!

What a world, what a world.

I suppose to try and filter that out would require looking to see how many new accounts received delegations of something other than 15 SP, though that would really carry only minimal information of any kind of use.

At some point, we just have to try and figure out what the most meaningful indicator is of useful activity. Myself, I lean toward active voting as one of the few things we can observe that might indicate some level of intentionality, given that most users of every social network are primarily consumers of content – but that being essentially indistinguishable from a bot purely interested in voting based on some sort of external bid or intent means that the information content is less than ideal.

It may be that the only useful information we can look forward to is next month's number of created accounts.

This is one of the strong reason that I keep advocating that we should be promoting "the steem blockchain" for its usefulness as the database backend for a set of entangled social media platforms and not as a crypto-commodity. Of course, to do so we would have to have developers who wanted to focus on the social media side rather than the wallet/commodity side, and I don't know if that is ever going to happen.

(No insult meant to the developers of Busy or Steempeak, the latter of which is my preferred interface, but they aren't really adding to the mechanisms around social media architectures. Hell, Gab, for all that Dissenter is a retread of several other comment systems created earlier last decade, is at least trying to do some new things. It's a shame that someone on the steem blockchain didn't get there first.)

lol I have to laugh, my turn to shoot you down was over due.

I do think the accounts being set up by other rather than steemit inc is interesting and as time goes this can be used as a filter on other analysis, such as churn. But for this report, for me, the % of accounts that complete their about section is a good but by no means accurate indicator of intention

Everybody gets five minutes in the autocannon. It is a functional inevitability.

But for this report, for me, the % of accounts that complete their about section is a good but by no means accurate indicator of intention

If I were writing or designing a system which created bot accounts and empowered them for vote harvesting, one of the things I would do for absolute sure is have it generate some sort of procedurally random account profile because it wouldn't have to be very convincing to be convincing enough. Pull a random landscape image off of Facebook or Google images for the background, generate some random demographic info, pull names from a pool created from Facebook and Twitter, and away you go. It would be very hard to distinguish between a legitimate new user and something of that order.

Which complicates the analysis, unfortunately.

One day I would really love to sit down and interview someone who runs one of the major bot farms and talk about the process that they use, how much automation they use or not, all of that stuff in creating and targeting bot swarms for social media platforms. I'm endlessly fascinated by that sort of thing and since I have a complete absence of moral judgment, don't mind talking to anybody.

It would just be interesting.

it would be very interesting

Coin Marketplace

STEEM 0.29
TRX 0.12
JST 0.033
BTC 62934.09
ETH 3118.65
USDT 1.00
SBD 3.85