User growth and retention

oskarth · July 13, 2020, 4:14am

Monday again, which means it is time for a weekly metrics update! Not much new here. Most of what was said here https://discuss.status.im/t/user-growth-and-retention/1782/12?u=oskarth still applies.

Two specific thing that have improved in last week:

@jakubgs setting up automated retention graphs here https://retention.status.im/
@cammellos confirming the two following things (see Growth metrics and retention - CodiMD):

peer id confirmed is the same across restarts
same across multiple IPs

I’ve also heard @andre is working on some cool stuff, but I’ll let him post that.

DAU

Up 12%, however hard to know if this is significant as it puts us back at two weeks ago baseline. If we keep it up for next week that’d be a good sign.

Retention

Unchanged (seems to be lagging a week, and only weekly for now). No reason to believe this would have changed.

arnetheduck · July 13, 2020, 6:54pm

leak? let’s add another sentence to make discuss happy.

andre · July 14, 2020, 9:00pm

friends!

TL;DR: Real-time cohorts dashboard lives here, and the referrals dashboard lives here

As you probably know, I’ve had my head deep in SQL queries, dashboards and cohorts. @jakubgs, @hester and I have been getting a reporting tool up and running that lets us see, as close to real time as humanly possible.

In doing so, we’ll be 100% transparent with regards to user retention, but also figures on the referral program that we’re about to launch, as I feel it’s important for everyone to see how their work shapes those numbers.

For now, the Cohorts dashboard lives here. It shows you weekly cohorts from beginning of the year (you’ll see it start in December 2019 as the first week of January didn’t start in 2020), and you can pick a date interval to see daily cohorts for.

I’m working on bringing in the referrals dashboard from a hosted instance, and that should get finished tomorrow. I’ll update this post to contain that link too.

In the meantime, I’d like it if folks thought a bit about improving these numbers. One angle that I think we can approach is to try and inspire people on cool ways they can use Status. There’s nothing more frustrating than having the tool that, for instance, Amazon whistleblowers need to reach out to some prominent engineers who are their voices, but don’t know exists. I also want to pair with @petty to write about how Facebook SDK caused an outage on over 20% of the App store, and what kind of data they collect and send about you when their SDK phones home.

Edit: I copied over the referrals dashboard too, which is now available here. This still points to our test environment, but will be changed into the production environment as soon as we have that in place.

oskarth · July 20, 2020, 2:54am

Happy Monday! Largely same as last week (and week before that) applies, so I’ll just do quick numbers update and diff.

Numbers

DAU

Retention

As you can see from the above graphs we regressed again, which points to this movement just being noise. It also means we haven’t really achieved our goal of growing on a week to week basis.

Other news

Andre and co put up some nice real time stats board for retention, referrals etc - see post above.
Andre and co are planning a fresh look at the roadmap and prioritization with retention and low hanging fruit front and centre. This will involve everyone who wants to participate to suggest and rank features based on (a) retention (b) effort axis. Details to come soon. (https://discuss.status.im/t/retention-focus-and-roadmaps/1826/2)

2 looks like the the very beginning of mindset shift mentioned in earlier post, so kudos to @andre for getting that ball rolling!

jonathan · July 23, 2020, 1:10am

In some recent discussions amongst marketing teams, we are struggling to define clear measurement plans for our campaigns and their effectiveness on acquisition and ultimately retention.

In short, the further down the marketing funnel one goes, from social media >> ultimately into Status, the less visibility we have into the journey. This is obviously by design to respect and protect data and privacy while living up to our principles.

What does this mean?

Well, unless we pause ALL marketing activity outside of one super specific initiative or campaign, we will not know with 100% accuracy which efforts are driving retention – and this is something we must simply live with and learn to deal with.

For example, there will always be organic comms across social, community channels, etc encouraging and educating people to try Status. There will be consistent evergreen CPA based campaigns across network, search, display, etc. There will always be word of mouth and the “network effect” that is outside of our control. All of these will hopefully lead to new, retained users.

So with that, we have taken a different approach to timebox specific initiatives and closely monitor and watch the impact on retention over a set period of time. Time based retention (based on weekly cohorts) instead of campaign based retention (based on tracking ID’s).

There is an upcoming timeboxed period of 3 week (July 27 - Aug 16) with the inclusion of:

1.5 release and comms
Brand initiative with hyper targeted content and promoted posts
Ambassador push
PR+ Launch in LatAm

This is already a lot but with information provided by channels such as twitter and playstore, we can have an idea as to what is most effective in driving installs. Once in app…it’s more or less a black box. So once again, have to look at time/date

Here is a measurement plan for the upcoming 3 weeks. With this, we should be able to see trends in impressions, referral sources, engagements on specific items, and thier impact on installs AND retention of weekly cohorts. It is crude for now. We will revise and improve after this initial campaign.

oskarth · July 24, 2020, 6:04am

Update: These updates have been going on for about a month now, and by now I think we are getting the gist of it. We are also starting to see signs of a change in mindset and focus, which is great to see. The challenge now is to keep up with it - regularly post visible updates to keep on top of it in terms of mindshare and work hard towards improving the bottom line on a weekly basis.

I’m going to take next week away from Status to focus on hacking Vac/Waku stuff, and after that I’m off on vacation. This means someone will have to take over posting updates, and I’m hoping someone like @andre (or @hester) will be able to pick up the torch, perhaps even with more intensity - paint the target red daily (chapter 5 of Swarmwise, https://falkvinge.net/files/2013/04/Swarmwise-2013-by-Rick-Falkvinge-v1.1-2013Sep01.pdf).

For people who haven’t seen it, there’s also another thread on Core retention, focus and a workshop on Monday focused on low effort wins for retention, do check it out: https://discuss.status.im/t/retention-focus-and-roadmaps/1826/19?u=oskarth

bmiller59 · August 23, 2020, 11:59pm

Hi, everyone. I appreciate this discussion and the challenges and opportunities outlined here. I have one proposal to throw into the ring that I do not see addressed in this forum yet that broadens your base of potential users, provides more reasons for people to install/onboard and possibly harmonizes the tension between crypto enthusiasts and privacy enthusiasts.

Become a Single Sign-On (SSO) provider using the new DID SIOP standard to enable Status to serve as safe and simple way to log into all of the websites I visit and never have to remember another password ever.

The technical jargon for this is making Status a “Self Issued OpenId Connect Provider (SIOP) for Decentralized Identities (DIDs).”

In other words, Status could become the preeminent decentralized, privacy protecting alternative to Google or Facebook sign on/log in. It could also replace password managers.

I have searched high and low and I have not found any existing DID SIOP providers integrated with crypto wallets yet. I offer this to you because you are fully open source and because I support your values and goals. This is an emerging space and one Status is well poised to capture and lead. This would drive various positive user growth and retention effects:

Improve visibility of Status by partnering with various existing websites and services that are frequented by crypto, privacy, open source and self-hosted enthusiasts. Why shouldn’t I be able to log into Discuss, Element/Matrix, NextCloud, WordPress, GitLab, Mastodon, Lemmy, etc. using Status? And these are just the obvious low hanging fruit. I already had to create two different Discuss logins just to participate in Status convos!
Keep Status launched, active and top of mind: Using Status possibly several times a day to log into different websites would achieve these benefits.
A chance to allow crypto to make people’s lives easier rather than more complex: I would love never to have to remember or store another password and Status could make that possible.

Some implementation opportunities to build momentum:

Create easy log in buttons and other components and publish and promote them with developers
Integrate this with with Embark to provider additional value add
Develop and market Status as an alternative to password managers
Create browser plugins that make it easy for Status to be used as log in on any site, even those that don’t natively support it
Offer Status as an alternative to serve as a two-factor secondary proof of identity instead of one time pads, cell phone numbers, etc for sites that are not willing to replace their primary log in method yet. You could create a developer package to help encourage this.

Here are some other references in addition to the official standard I linked above:

BTW, thanks for all your work on Status! I am very aligned with your goals and looking forward to getting more plugged in over time. And to using Status to log into all of my accounts.

hester · August 24, 2020, 7:57am

Thanks for starting the conversation on this @bmiller59 DID’s definitely deserve more attention