Alien Shepherd Post-Mortem

Table of Contents


I released my first game Alien Shepherd on the 25th of February on steam.

Alien Shepherd is a precision platformer with a flying mechanic that allows you to go in any direction for a short duration. There's two different biomes and 50+ levels of increasing difficulty.

I started development in December 2020, and worked on it after my full-time job until the release. My sister worked on the graphics and I worked on the development.

Because I have a lot of experience with Python, I started working on the game using pygame. But, I had a lot of difficulties making builds for Windows and making shaders work, so I switched to Heaps.io after about 2 months. I didn't want to use an engine (like Unity), I like writing code, so a library made more sense.

The initial goal was to release a game on Steam, and have fun doing it.

Marketing & Performances

At launch we had 80 wishlists. The only marketing we did was creating a Twitter account and sending some tweets, the #WishlistWednesday had the best conversion from tweet to wishlist. Right now we're at 115 wishlists. The release itself brought more wishlists than I anticipated. I guess the new games appear on a Steam page somewhere and that brings some traffic.

Median time played is 47 minutes, which is lower than Steam average, but I expect this to go up because the game has only been out for 36h.

Since release, we have sold 11 copies of the game for a gross revenue of $46 and a net revenue of $36.

Most copies were sold in my home country (France) even though I marketed the game in English. This tells me that people who bought the game are friends of friends and not people that came from Twitter.

What worked

There are a lot of things that helped releasing the game:

  • Having a clear deadline in mind from the beginning (I wanted to spend 1 year on the game)
  • Not hesitating to reduce the scope (we could still be working on it if we implemented everything initially planned)
  • Buying the music ($14) (we aren't musician, I tried to make some music but it took so much time for very bad results)
  • Making sure I was having fun developing the game (I didn't want a second job, I wanted a hobby)
  • Working with someone as motivated as me (I wouldn't have released the game alone)
  • Using the level editor LDtk (made by the creator of Dead Cells)

What I would do differently next time

For this game, the goal was to release a game I could consider "finished". If I ever attempt to make another one, I would aim for financial success. That means >$20/h return on time spent developing the game.

To achieve that, I would start the marketing much earlier in the development process. I would focus on a vertical slice of the game and polish it until I get enough wishlists. Only with enough wishlists, I would start spending time on creating content.

I would also not create a platformer. They are, by far, not the best performing games on steam. And I feel like the ratio between play-time / development-time isn't high compared to something with high replayability like roguelikes. The level-design was definitely not what was the most fun for me.

Date: 2022-02-26 Sat 00:00

Emacs 27.1 (Org mode 9.3)