Viridi on Mobile (One Month In): Curiosities!

Hi internet! I decided to write this update after a couple of oddities cropped up around the mobile release of our succulent simulator Viridi. This is our first mobile game, so it’s possible these things that seem odd to me are actually old news to the rest of the world. Either way someone learns something - either you or me. Apologies for the disconnected mess of thoughts and data ahead.

Boring generic overview

In a little over a month, Viridi has moved ~150k units on the Android Play Store, and ~100k on the iOS App Store. The pickup rate seems to be unsteady but accelerating on average. Here’s our ‘current installs’ curve on Android, for instance. (Hoping that last plateau bumps back up soon, but who knows, all is chaos etc.)

We just went to Casual Connect, where I was exposed to a lot of mobile jargon, and thanks to Unity Analytics, I can actually look up some of these various stats. For the group of people who know and care about these, our ARPPU is a little over $3, our ARPDAU is a little under $0.05, our retention rates (are changing a lot (mostly going up) but) are currently around 55% on day 1, 20% on day 15, and our conversion rates are ~1% 15 days from install. I assume these numbers are low, since we’re not designing for any of this stuff, but I would be curious to know better.

Curiosity 1: Tutorial = lower review scores

Reading our more negative reviews, we saw a trend where people would mention that the app didn’t seem to have anything to do, and was entirely based on IAP. This is sort of true, since the point of the app is actually the user doing very little: just focusing, waiting, paying attention, and slowly nurturing their plants. But we thought, “Hey, a lot of these users are just bumping into us on the store, and maybe don’t know what the app is. They’re probably coming in for their first experience, seeing plants that don’t do anything except die when you overwater them, and then uninstalling the app out of frustration. We probably just need to be more clear in our messaging.”

We decided to add a few dozen words of tutorialization. Very lightweight, just enough text to signal to the user that the game also happens while you’re away, and that closing the app is part of the experience. We looked at Neko Atsume’s tutorial as inspiration, but decided to take a lighter approach, asking the users to close the app and come back later, but not forcing them out. I was excited to see that the Google Developer backend supports A/B split testing. I wanted to be able to see with certainty that the revised introductory experience was leading to happier players.

Above are the results. Build 18 was the build with no introductory tutorial, build 20 with. Those numbers are stars out of 5. There is indeed a strong trend, but it’s the opposite of what we expected. We shouldn’t have second-guessed ourselves maybe, but then we never would have seen this interesting data point. Though it’s hard to know what exactly to take from it.

Curiosity 2: 90% of our iOS installs are in China

Presented for your consideration: Our installs yesterday on iOS:

We’re not even localized in any way. And our installs by territory:

Additionally on iOS, we had 14 million impressions in China yesterday, up from 50k worldwide the day before. Not sure how that happened. Please do let me know if you have some ideas.

Curiosity 3: US market seems much better on Android

Here’s our installs by territory on Android:

We have ~5k iOS installs in the US, versus nearly 50k on Android. This seems to mostly be based on organic factors. Maybe the Play Store has significantly better discovery algorithms? We were consistently told that iOS was the more important platform to be on. So far, that doesn’t seem to be the case. Though with our growth in China, maybe so?

--

Please do provide any thoughts in the comments! But, you know, be nice.

[cross-posted from http://www.icewatergames.com/blog/]