Holding On To Our Newly Onboarded Users
Hive has a user retention problem. What's worse is that we struggle to retain users even while our on-boarding in the first place is extremely low, so of the few users we manage to bring in every day, even fewer stick with Hive. Today I would like to examine our retention rates for the newest users and to use profile location data to see if there are some places where we are doing better than others.
In yesterday's post I took a look at how user retention compares on Hive for older vs. younger cohorts of users (a cohort is a group of users who joined around the same time). In that post I shared the chart below, which shows the percentage rate of brand new users each month who continue to comment into the second month.
Steem era average: 36.57%
Hive era average: 37.97%
As you can see, this metric has been very poor for the lifetime of Steem and Hive. Usually less than 40% of newly active users make another comment in their second month. This also doesn't seem to have improved in the days of Hive vs. Steem, if there has been any improvement it has been marginal.
That said, perhaps the chart above is too influenced by bots. On Hive and Steem, account automation typically focuses on votes and comments not full posts, so we can reduce their influence on the stats by restricting the graph to just accounts that make posts. The newer chart also shows some longer term moving averages.
Steem era average w/o comments: 35.74%
Hive era average w/o comments: 37.81%
By looking only at posts, retention in the early days of Steem gets slightly worse, but it's pretty much unchanged for Hive. The moving averages also show that the rate of improvement in brand new user retention is very slow.
Retention by Country
I have already shown how some national groups have better retention overall, such as Venezuelan authors consistently staying on to post again each month at rates of 75% to 80%. However even that group is dominated by older cohorts, so how do they fare with new user retention?
Well first we need to set a control. If we are looking at location data, we know that only 11% of users ever set location data in their profile. Of those 82% have set their location within a month of joining, so there should not be a huge difference between more recent data and older data in this regard. We can see below how merely setting any location data at all makes you more likely to stick around than the average new user.
For all users, the average retention rate since the beginning of Hive is about 37.88%, but that improves to 51.14% for users that set location data. Simply applying a location in your profile increases the likelihood that you will stay for a second month on Hive by 35%.
Now that we have a control with which to compare, let's see how new user retention looks for some of our largest national groups.
Since the split, brand new users with a Venezuelan location have had a retention rate of 62%. This is the best I have come across, and their better retention is clearly seen on the chart. We can estimate that including all brand new Venezuelan users, the adjusted retention rate is probably about 46%.
Let's look at the US for comparison.
While the chart is a bit all over the place, due to low numbers at some times, the average is a mere 41.72%. When we take account of the fact that merely having location data increases your retention rate, brand new US users may have as low an adjusted retention rate as 31%. This drags our overall retention down.
As an aside, perhaps this in part explains why Hive has little traction with crypto Twitter, which is fairly US-centric.
Here are some of the rest of our largest national groups, all with higher than average brand new user retention.
Cuba average: 60% or 45% adjusted.
Philippines average: 55% or 41% adjusted.
Nigeria Average: 55% or 41% adjusted.
UK average: 58% or 43% adjusted.
Conclusion
Our brand new user retention is poor across the board, however in some places it is better than others, and those are where we are likely to see some growth. As Venezuela starts to represent an ever larger portion of the Hive userbase, our new user retention may improve, but even that will be marginal. Substantial improvements will depend on identifying the reasons for our low retention rates and resolving those problems.
My statistics and analysis posts take many hours each to research, chart and write, so if you find them valuable and of interest to other Hivers, I appreciate your support in sharing, commenting, and/or upvoting my work. If you're interested in these kinds of stats posts, click the 'follow' button on my profile, or subscribe to the Hive Statistics Community which features daily Hive stats posts from @arcange as well as less regular posts from myself and others.