How to Engage and Retain Users ( Backed up with Steemit Data)

in #witness-category6 years ago

Steemit is an experimental decentralized blogging/social platform created on blockchain technology that allows users reward content creators by way of up votes which are attached with a value.

At Just over 2 years old and with almost 1m registered accounts, because of the decentralised nature of Steemit and the self-governance of politics and business ethic’s, Steemit has become many things to many people.

For some it is a blogging platform, where authors publish content which they hope adds value and is rewarded financially.  For some it’s a social place to share their lives and interact with other, with the added benefit of possible rewards.  For others it’s a place to do business, earn a return on investment, or use their development skills to create awesome apps and projects.

At first Steemit was rather confined to developers and those in the know around the crypto space, but as Steemit becomes more mainstream and the user base grows, the niches that Steemit caters for have and will widen further.

At the heart of Steemit making sure everything works are the Steemit Witnesses.  These people devote their time to ensure we all have a stable and secure platform.  Many of them also create apps and tools offering different user experiences or tools.   As I said, Steemit is decentralised.  There is no centralised organisation running the platform.  Without the efforts of these people there would be no Steemit.

In addition to running a stable secure network the witnesses also have a say in future changes to how the blockchain works.  Any changes made to the code and how steemit works, must be agreed by a majority of these witnesses.

For Example, if someone wanted to change how the rewards are calculated, this would require a change in the code, and an agreement between witnesses to run the new code. Not one person or one organisation can make an implement these changes on their own.  

So these witnesses, a curtail part of the success of steemit are working away providing this stable network and awesome apps.  If we were to compare this to a centralised organisation, this would be the IT and Systems Development and Operations Departments.  A department that is served well under a decentralised platform.

I know some of you will say, well why are you comparing this to a centralised organisation?  Well that’s easy really. At the end of the day, 99% of the people on steemit have a common goal.  Profit. Same as most centralised organisations.  And therefore, like most centralised organisations, other departments are needed to.

And wow is this happening on Steemit.  How awesome it is that random people are starting departments on Steemit.  For Example, on marketing we have promo-steemit and for business intelligence we have blochchainbi both established tags and communities on Steemit. We have all these side communities working on the ground and active within steemit.  It’s all decentralised, yet organised.

Lets get back to the witnesses again.  So we have an awesome bunch representing the IT department as witnesses. But what about the other departments?  The people on the ground.  How are these represented within the witnesses? 

Well there are also some witnesses that are working on the ground helping new steemain understand the platform and grow their accounts.  Running other departments.  Meet @abh12345, half of the steemit witness team @steemcommunity .

@abh12345 has been working on the ground encouraging manual curation and engagement though his Curation and Engagement League.  The ideas rather simple.  There are two leagues, a league that rewards those for maximising their curation rewards with the SP their have and the Engagement league (or league of excellence) which takes in a number of actives on the block and weights these actives producing a weekly leader board. (We could call this department Community Growth and Account Acceleration, it could also fall under the umbrella of on boarding)

Let’s take a look at some data

 

The data above looks a bit empty right? Let me tell you why.  Since Steemit Started 995,824 accounts have been registered on Steemit.  Of which only 133,956 (13.45%) have been active via votes or posts/comments in the last 14 days.  That leaves 861,868 non active accounts in the last 14 days.  These figures are shown on the top of the visualization above.

Below this the first rather blank looking charts shows at what day people drop off Steemit.  Hard as it is to see, of the 861,862 accounts that have not been active in the last 14 days,  +500K accounts which is just 59% have not posted or voted since the first day they registered on steemit.

Based on this the median drop of day on Steemit is 1 and the average is 31.65

Lets put some filters on this data set.  As a majority of accounts that registered on steemit never make it past the first day, we can completely exclude these from our analysis

   

Now at least we can see some of the data.  By excluding all accounts that have not been active since their first day we are left with 487,661 accounts of which 72.53% have not been active in the last 14 days.

 The median drop of day is now 32 and the average is 75.74.  These numbers are really improving however the distribution of drop offs by the day in which they drop off is still very much skewed to the right.

Using this as a base we can now compare Steemit overall to participants in Asher’s leagues.

     

Notice a difference?

Of the 357 participants in the league 93.28% have been active in the last 14 days and only 24 have dropped off.  For the accounts the dropped off, the median drop off day is 81, with an average of almost 104.

Okay so some of you might have spotted this but the sample population taken here is all of Steemit and all of the participants in the league.  By taking the full population the data is slightly skewed because one could argue that the people that joined Asher’s league when it started were already established on Steemit.  

So what if we add some more filters to the data.  This time only looking at Accounts set up since the beginning of 2018.

  

Leaving on the filter to exclude accounts that have not been active since their first day, we are left with 222,971 accounts registered this year.  Of this 40.96% have been active on the last 14 days.  For  the accounts that did drop off, the median drop off day is 18 and the average is 27.46.

Now let’s look at participants in Ashers league

 

From this we can see that 107 accounts registered in 2018 and joined Ashers league.  Of this almost 87% of accounts have remained active.  That’s more than twice that of Steemit in general (and they are all human on Ashers league)

For the accounts that did drop off, the median drop of day is 46.  That means that the people that do leave, don’t lose their interest as fast if they are in the league.  In fact they stay over 2.5 time longer than people that do not participate.

So what’s the moral of this story (apart from saying Ashers league rocks)?

Retention Rates on Steemit are low and churn rates are high.  This can be backed up by the following post

https://steemit.com/utopian-io/@paulag/steemit-churn-rates-q1-2018

Witnesses not only have a role in providing a stable secure platform and a say in changes, but they are also responsible for the continued marketing, on boarding, and growth of Steemit.  They need to engage and retain users.  We can’t expect all witnesses to be all things.  But we can ensure that each department on Steemit that provides value to the continued development of the platform is represented with in the top witnesses.

Asher’s league has become very successful.  All you have to do is take a look at the posts and see the level of engagement via comments to see that.  And the league has spread.  We now have @girolamomarotta doing a league for Italy and I run a league for Indonesia.  Me, being second half of @steemcommunity witness team.

Have a spare witness vote, or looking to make changes to your witness list?  Then I do hope you will consider voting for @steemcommunity

All data for this post was taken from Steemsql.


Sort:  

Well well, look what we have here - some data to back up what I, and others in my leagues, had suspected! :D

Thank you @paulag for grafting this one out!

Of this almost 87% of accounts have remained active. That’s more than twice that of Steemit in general.

BOOM!


If community members would like to run a league, we can generate the data for you community and just need a list of names.

eg. 'name','name2','name3'

You will need a bit of basic Excel knowledge and that's about it!

We can also run this report to track the success of other communities making a difference to user on the platform. It might be worth comparing a few different initiatives. Especially those in receipt of funding

Or perhaps to track a tag? #meme ? #eos ?

I'd be very interested to see this. Two communities I am involved with use a specific tag #teamaustralia and #steemsilvergold. If we could agree on a benchmark for tag usage I think it could be quite valuable in helping identifying what communities are thriving/growing and which one's aren't. Then we could have a close look at what they are doing right (or wrong)

Great. @paulag - Our first customers. (And they have precious metal :) )

Happy to work with you on this if you think it would be of use. Cheers.

Yes, I need to figure out how and when we can do this. It'll need a bit of work so I'll have to block out some time this week perhaps.

Yeah no rush, depending on what's required I've probably got some scripts that can be adjusted to suit.

So, it's been a couple of days. Is this being worked on? Just wandering. :)

not yet @glenalbrethsen. Sorry, been so busy answering comments i havent got to data. Just added it to my list. I will hit you up later to fine tune the details you need :-)

I'm interested in the results, but I think it's @buggedout that has the specifics this go around. I was kind of answering off of yours and what @abh112345 was saying about tracking initiatives and tags, but didn't have anything specific in mind. Was mainly curious to see how these others might be doing in relation to the leagues as far as participation and retention go. So no hurry and no worries on my account. Just interested in whatever results there may be. :)

I had sent to a message on discord, i will message @buggedout.

I'm leaving this one to Paula, she's got the scripts in place! :D

Good work as always.

Superb job you both have done!
Asher with your league. And Paula both with the local Indonesian league as well as this analysis.

I think we should find supporters for a new league for all newbies having an account that's less than 3-6 months old.
And build a bot to send a comment to all new accounts.

no comment bots lol, till then I was loving your comment.

Why not?
In this case it's for a good cause....

everyone justifies their reasons for bots, good bots, bad bots, comment bots, voting bots. where is the human element of steemit, because it wont be found in lines of code

I totally agree with you. Preferably we would ban all bots. But in the past weeks I've learned that it is impossible to get rid of those annoying and sometimes destructive bots. So I thought let's do like everyone else, join them as you can't beat them.

Congrats on this proof from @paulag... you are right that we knew, but its gonna make it much easier to explain to those that aren't engagers on their own ;)

Interesting figures @paulag. I hear a number pf witnesses saying how wonderful it is that we are nearly at 1 million accounts. I don't see that it's much to shout home about when churn and spam accounts are so high.

I'm glad to see the @abh12345's league is attracting people who are truly interested in engagement. Having a 200 SP delegation from him was a big help to me in learning how to effectively spread my vote around.

Good that it's spreading to other countries too. 😊

I don't see that it's much to shout home about when churn and spam accounts are so high.

Yeah, the total of individual accounts is probably more like 10% of the grand total.

I have 7 accounts, bernie has 100+, ned has a load, dan has a load, we have 100's of bots, 100's of community accounts, etc, etc.

Shilling Steem/Steemit for its number of accounts isn't the right approach imo - perhaps if we have account based voting which I'm not sure is planned 'here' in the near future.

I'm not sure account based voting would work here any better here anyway since the more power you have the more votes you seem to get. At least that's been my observation on accounts where people have invested and suddenly have higher SP. Their content is still crap but they get even more votes. 😂

Hopefully, each account would have one vote, and each account would be verified to a real person.

You've lost me on the one vote bit. 😕

One vote of equal 'value' - So Trending for example, could be order by number of votes by unique people - that would surely move the genuinely interesting stuff to the top?!

Asher has been one of my favorite steemians since I returned to the platform. You are both pretty awesome. <3

This looks like up-vote material to me! haha, aww thank you! :D :D :D

No, YOU'RE upvote material!!!

Are you using steemlookup yet? It is curator gold.

blush

I have checked it out yeah and it's improving all the time. With access to the blockchain data though, I have some queries to do similar that help me out.

Not all of us are awesome enough to make their own queries to the blockchain!!!

I joined here in the end of January and following @abh12345 for very long time since then. I heard about his curation league but didn't understand anything. It looks awesome and I think I should participate in it if he conducts it again.

Indeed witness voting is a crucial thing everyone should give a thought upon. Are you a witness too?

Myself and Paula run the witness @steemcommunity.

I conduct the contest weekly, and I will add your name :)

Thank you brother!

+500K accounts which is just 59% have not posted or voted since the first day they registered on steemit.

Is there a way to tell how many of those are accounts that were created but never used? I have a feeling there's a lot of "churn" from users who sign up and then either wander off during the enrollment lag or find a non-free way to get an account.

Could you run a graph with number of SP in inactive accounts? I imagine there's a cliff and I'm curious where it is.

Is there a way to tell how many of those are accounts that were created but never used?

Yes I think so, but the 'wandering off during enrollment' lag is something I don't think we can investigate from the Blockchain. That would be interesting though, and Steemit inc will likely have some numbers they could run.

there are a lot of high values accounts that dont use their SP, are you just looking for the accounts that started and never did anything?

I'm curious about that, but what I'd really like to know is if there's an inflection point in SP where a user becomes far less likely to leave. If we can say "I found a quality poster let's get them to 200 SP quickly and they'll almost certainly stay" then that becomes something we can build a project around.

I disagree that Witnesses are decentralized. They are "trusted" peers and their number is low.

If someone really wanted to, they could compromise each one and direct the network however they wanted.

Its a far cry from larger networks like Bitcoin - which has resisted social-engineering and technical attacks.

The jury is still out on Delegated-Proof-of-Stake, as it seems to copy a lot of the flawed rules that the legacy fiat system does -- namely, the more tokens you have, the bigger say.

The retention rate should be a huge red flag for the Steemit developers. They recently said they added some bot-combating measures, but I am not holding my breath -- I deal with results, not hopeful wishing.

Fair points.

I'm hoping to see SMT try and combat the ' the more tokens you have, the bigger say', particular with regards to ordering content here, and elsewhere if/when SMT are out in the wild.

One account one vote 'account based' voting, should probably be the default sorting order for interesting content, and Trending can then be renamed to 'Adverts'.

Cheers.

Current Retention should be a warning flag but its not. Pushing ahead with development and improvements on the block are the proirity, or thats how it feels to me.

I believe steemit is way more decentraised than it was before. Just look at the changes in the top witnesses in the last 12 month. Yes they are a 'trusted few, and by the nature of few makes it impossible to be fully decentralised. Maybe to improve this the few needs to be increased to a few more. I dont see many of the existing few currently willing to do that either.

Other than the huge amounts of bots, I personally know some people lost their passwords to their accounts (very unfortunate,) so I'm pretty sure there around %30 of steemit accounts can't be used anymore... yeah... There's that.

30% would be rather high but would account for a large portion of the inactivity. Interesting.

Well, it's just a speculation though. considering the samples I stumbled into (posts about losing a previous accounts, people registering spare accounts to try different things, bots that no longer operate... and such.)

you have given me an idea for a data analysis post. Pity i dont have the time to do it this week, but I think I will try search all posts on steemit and find as many that I can that have lost their keys. Awesome. Thank you!

Wow!! Good idea, if you remembered contact me (just comment on any of my posts) when you post it.

This is a very good post and you are right

To listen to the audio version of this article click on the play image.

Brought to you by @tts. If you find it useful please consider upvote this reply.

Somebody needs to develop an on-boarding process for new accounts. The huge numbers that drop off after one day are probably seeing the giant mish-mash of content with no effective way to filter and sort to find what they want.

So the easiest thing to do is chuck it in the bin.

You are so right. But its hard to convince the people that need convincing that this should be as much of a priority as new apps or tools

Coin Marketplace

STEEM 0.28
TRX 0.11
JST 0.034
BTC 66518.22
ETH 3187.49
USDT 1.00
SBD 4.05