What We Wish We’d Known About Audio Games Design (Part III)
Our game StarLanes was the first massively multiplayer online game for Voice. It also has, as noted in the previous post, a complex, complicated codebase. StarLanes is so complex that it eventually fell victim to what we call “Voice Interface Cluttering.” Our solution was to break the StarLanes game up into four separate but interconnected mini games. Events occurring in one game impact events in the others. Various systems would keep the games in sync.
But before we could execute this game quartet plan, we hit another problem. It had to do with user acquisition.
The Alexa Skill Store contains well over 10,000 applications at that point. It is massive enough that users no longer “stumble across” our game randomly as they once did. The Alexa store ranking system favors applications with small numbers of universally high ratings, and there isn’t a big enough audio assistant install base for word to spread organically. As our games have grown less discoverable we saw the expected drop offs in new users.
We decided we would experiment with advertising to make up for the shortfall. After all, we had just begun monetization of StarLanes skill (selling T-shirts via Merch with Amazon) which gave us a small budget for marketing.
We started with Facebook ads and Google Ad Words as we felt their keyword driven approach would allow us to hone in on potential and former players. We were looking for Alexa owners who also liked strategy gamers. They also had to be curious enough to have explored games on Alexa. We tied our ad buy to our budget, setting up a few ads designed to draw the users in.
Except that they didn’t.
We had a pretty good conversation rate, i.e. higher than we were lead to expect for this sort of advertising. Unfortunately, the overall numbers were very, very low. We never even reached our budget numbers. Very few people used Google search to find audio applications. And there weren’t any affiliated sites with the right userbase for Google to match up with what we wanted.
In addition to that, for the ads we did serve up, we were unable to measure any users coming from them. We tried other advertising, one at a time, and measured the results. Nothing moved the needle above the background noise.
Fortuitously we did manage to get skills listed in the Alexa Companion App’s sliders, and in the Amazon weekly circular. These did provide a measurable boost to the new users coming in. But once the promotion stopped, so did the new users. And, since Amazon do not currently allow you to buy advertising in these slots, such promotion was only available at Amazon’s discretion.
The take home lesson is that this is really hard. We have not yet found a good solution to this and it’s one of the two main barriers we face today to producing successful games on these platforms.
* * *
But it gets worse. User retention is also hard.
The Voice platforms haven’t (yet) come up with means for creating bookmarks, or favorites or even breadcrumbs. We as developers have to rely on users to remember the names of our apps and how to ask for them. We as designers haven’t yet come up with sensible solutions for forgetfulness.
There’s nothing more soul destroying to see a vibrant, energized fan base slowly slip through your fingers.
People come to play games. They might find your game through the hoopla of the initial release or maybe they found you through a fortuitous promotion or a handily timed winter break. New users will appear so long as the promotion continues. They will play the game for a while, and then they will move on. That’s part of game development. The aim is to create enough content or engagement to keep them on the game long enough to make a return from them, in whatever form you seek. But Voice has a very steep rate of decay.
Looking into this, we found a number of structural problems.
- You can’t create a book mark for a voice app.
- You can’t put an icon on a desktop.
- The platforms now offer ways to create a URL that links to the web directory for the game. But as with advertising, browsing the web is so far removed from using Voice, that the transfer rate is very, very low.
- There are no breadcrumbs.
- There are no Favorites or Recent lists.
- The platforms do not enforce naming or gaming exclusivity.
You, as a developer, have very few options to reminding players how much they like your game.
There are a few ways to get a user’s e-mail address and send a “we haven’t seen you in a while” reminder. But these involve an up-front opt-in that can be off-putting, which will reduce your already low acquisition rate. And, you will need to revise your expectations downward when considering appeals through the web. It’s like asking someone to take the train and then a bus to come to dinner. The transfer rate from an e-mail to a voice invocation is very low.
The take home lesson is that, this too, is very hard and represents the second main barrier for the platform. The only solution we’ve found is to create games so compelling and engaging that they become one of the user’s sole reason for using their device. But that brings us to the next problem.