Ellipsis is an award-winning action puzzler with retro-styled visuals and absolutely zero text. Designed from the ground up for touch devices, it was released on iOS in February 2016 and for Android on June 16th. Its intuitive concept is easy to understand but you soon uncover a deep and challenging universe of ever-evolving levels. Ellipsis is a polished game experience, developed with a bit of idealistic, opinionated approach. Released as a premium game, it features no text, no ads, no IAP. The developers headed by Yacine Salmi explore how these decisions impacted our development and release.)
As a small indie team, we had a very limited marketing budget. We promoted our game heavily but it is a challenge to stand out from the crowd of other great games out there. We focused on the things we could control: first and foremost, making a great game. We invested two years into making a simple concept as fun as possible. We also continuously spread the word about Ellipsis throughout development. Marketing is not something you can simply start at the end.
Background and Concept
Ellipsis was conceived at a game jam, exploring the possibilities of direct touch controls. Stefan and I regularly jammed on game ideas ever since we met at a design event in 2012. This time we wanted to explore what was possible if you went further than the usual taps and swipes and directly mapped the finger to an object. This fun prototype quickly morphed into an abstract game about touch control and swift evasion.
Development continued sporadically over the next months until we could marshal the resources to take a full plunge. Over time and through constant iteration, the game evolved from a basic experience to a complex universe of over 130 unique, diverse and ever-evolving levels. Code, design and art responsibilities are shared between us two and audio is handled by our associate composer Filippo.
Our goal was to throw the player into an abstract universe that feels alive and invite them to explore it. We took a very idealistic approach deciding not to use any text. This posed a lot of challenges to get right. We also decided against ads and in-app purchases. We took a risk by releasing the game as a premium title instead of F2P but it’s the only model that really made sense aesthetically for Ellipsis.
We developed Ellipsis over a period of 2 years, and finally released on iOS in February. Since then, we haven’t stopped improving the game, having recently released a 1.1 update for iOS, launching Ellipsis on Apple TV and now the big release on Android. We will finally have a bit of time to catch up on sleep before we move on the next project.
Telling a Story of a Journey in an Abstract Setting
Ellipsis is a fun game of reaction and skill, with abstract enemies and environments. We also wanted to tell a story with Ellipsis. It was important to convey the journey of the player as they explore this vast world.
Having a story was also critical for us, internally. It helped guide our decisions when designing new levels or adjusting the progression. It became another constraint, helping us introduce new elements, enemies and features in a consistent and coherent manner.
Not every player will grasp the story we are telling, we accept that. It is not even critical that they do. It does not affect the fun of the game. The story simply adds another layer of depth for those players that do notice and appreciate it. For us, it is enough that the story exists and that the feeling of a consistent universe is present. The abstract shapes and colors give the player the freedom to project their own interpretations.
Leading the Player Without Text
A major challenge was how to teach the gameplay mechanics to the player without the use of text. The minimalist style made it challenging to introduce certain elements. It was especially difficult for us to explain the overview map that allows the player to navigate between levels and puts the single levels into a spatial context.
We did a lot of things to address these challenges. First and foremost, we playtested repeatedly and with new players. The first few levels of the game are key. Every time the player becomes frustrated or experiences the wrong feedback (something completely opposite to their expectations) – we fail. Each failure has been an opportunity to improve the experience.
We iterated endlessly on the initial levels. Through a series of playtests and showcases, we made sure we had access to a constant stream of new players. Once a player “gets” how the game works, they would become almost useless for testing the tutorial. We would simply hand the tablet to players, telling them to “explore” and no more, observing how they reacted, experimented and learned.
Consistent use of color and shapes was very important. For example, using colors such as white (neutral), red (deadly, pain), purple (patrol behaviour), green (speed) and blue (player, good) – we could give hints on what to expect from the different objects and enemy types in this universe. We worked hard to stay consistent.
We also unlocked features progressively, revealing new elements of the UI piece by piece. The map only appears after the first level. The map camera is only unlocked after the 4th level. Omitting these elements in the first levels really helped guide the player. By release, we reached a state where most players understood the game within a few minutes without any explanations necessary (and with minimal frustration).
Releasing the Game and Marketing
Completing the development process and creating a good game are only two parts to the puzzle. In order to get noticed, you have to work really hard, building awareness, contacting press, creating stories for your game. The market is more crowded than ever, and a good game will rarely become successful just because it’s good. Since we released without a publisher we had to manage all press contacts ourselves for which you should really reserve a lot of time.
The First Month
Here are the sales numbers and what caused which sales spikes:
We got active at the TouchArcade forums, reached out to 150+ journalists, 30+ YouTubers and attended several game events before releasing. Luckily, we got a few smaller reviews right at release. Most importantly, though, we got featured in “Best new apps” and “Best minimalist apps” in the App Store on several continents (The high sales in the first week are mostly due to the Apple Feature). Unfortunately, the sales went significantly down after the first week.
It is really important to be persistent with press contacts and leverage previous success. For example, we did not get much press coverage at release. However, when we sent out a second mail to press mentioning the positive reviews on AppAdvice and TouchArcade as well as the Apple feature, we got the attention of bigger websites.
So far the most important one has been a review on Spiegel Online, one of the biggest German newspapers (check out the spike around Feb 25th). We almost got our highest number of sales, and that by a sales boost in Germany only. Consequently, we reached #1 Puzzle Game, #2 Action Game and #4 Paid Apps in the German charts.
Unfortunately, Spiegel Online was the only article that really boosted our sales. We contacted a lot of people, got featured at one of the biggest Austrian radio channels (FM4) and got several articles in different newspapers, but the impact on sales was quite low.
A third smaller boost was on March 4th when we got featured by Gameranx in the Best iOS & Android games of February 2016.
It is also nice to see websites you haven’t contacted talk about your game. This 18-minute-long Japanese lets play video ranks as one of our favorite coverage so far:
Without text, we could immediately release worldwide without worrying about localization. Or so we thought. In order to have a better shot at getting featured, we still translated the App Store description in 24 languages. As an experiment, we translated to Swedish but not Norwegian. Both being largely English speaking countries, we wanted to measure the impact of localization. We had 50% more installs in Sweden, but as our overall sample size is still quite small (relatively speaking), we don’t feel comfortable drawing any larger conclusions. The difference in install size could easily be a statistical error.
Releasing a game with no text led to some unexpected marketing challenges. Aside from still needing to translate app store descriptions (and keywords), we found the process of describing our game extremely challenging and spent a lot more time than expected crafting our message and description. Describing and selling an abstract game is quite challenging and we’ve gone through many iterations. Even now we continue to refine it.
What We Learned About the App Store
Releasing for the App Store was an incredible learning experience. Here are some of the things we found out along the way.
Every Thursday the store features page gets a refresh, so releasing on a Wednesday afternoon/evening is a generally good timing. With recent news it seems like the App Store might get a more frequent refresh, but the overall weekly refreshes are also likely to be there. Make sure to have the build approved at least a week before, and preferably 2-3 weeks earlier. This gives you enough time to contact Apple and inform them about the upcoming release. This also gave us time to send out promo codes to journalists who love being able to test a final version before release.
As soon as we pressed the Publish button it took roughly 10 minutes for the game to appear in the US App Store and about 20 minutes for other countries. This was surprisingly fast. Submission approval times have dropped massively since February. Previously at 7 days duration, our last submission in May took only 2 days to get approved. It seems that roughly 90% of releases currently get approved within 48 hours.
We spent the first few days after release continuously switching between stores to see if/how we got featured (we later learned that AppAnnie is an easier way to keep track of such things). Apple group features roughly by major region (North America, Europe, Asia, South America), but with variations across some individual countries. Some stores are also curated much better than others. For example, individual categories in Europe (Puzzle, Action) are filled with spam whereas in North America they are well-curated. But this is also constantly changing as Apple continues to improve and try various different promotion strategies.
Apple also creates small collections of themed games, some of them appearing only in specific regions (cat games in Japan, special collection for Chinese New Year), others showing up in a lot of countries (we were super-proud to be featured in “Best Minimalist Games” besides some of our favourites like Blek or Monument Valley). Popular games really appear in a lot of different collections. These collections don’t seem to lead to huge sales however, unless they are prominently displayed.
What We Could Have Done Better
Despite repeated warnings from fellow devs, we managed to launch with a major crash, albeit only for older devices. This is especially frustrating for us because we took great pride in the stability and polish of our game. The game is solid, but through an error in oversight, we shipped a game supporting iOS 5.1 and higher without even testing on devices older than iOS 8. Complete fail. We simply didn’t have a device around. Basically, because it ran solidly on an iPhone 4S and iPad Mini, we just assumed it would work fine on other devices.
And even though we ran a TestFlight beta with 100 testers, we still didn’t catch this…because the irony of the whole situation is that only players with iOS 8 or higher can even use TestFlight.
That was mistake #1. Mistake #2 was shipping without symbols. No symbols meant we couldn’t obtain crash details that you can normally access through Apple’s voluntary feedback system. We removed the symbols in the mistaken assumption it would help us minimize the space used by the app. As of this writing, we’re pretty sure the symbols stay on the Apple servers, so there wouldn’t even be a size gain. Sigh. Could you please provide a screenshot here? Sounds like something I don’t know about but interesting, would be cool to add a visual to understand.
Turns out that aside from the iPad 1 (the only device that doesn’t support iOS 9), some people don’t update past iOS 7 due to compatibility issues with some apps that are no longer well supported.
To round out the irony of the situation, the crash itself was simply hard-linking to the iCloud library. Making it optional fixes the game for all iOS 6 and 7 users. It was a two minutes fix.
Luckily we have addressed these issues in our most recent update (along with other major improvements, many new user-requested features (achievements, game center, replay kit, screenshot sharing), polish on the levels and a free level pack. All for free of course.
We had a good start but, unfortunately, the number of sales dropped after the first few weeks. Still we push ahead. We just released a major update on iOS and launched on Apple TV. We were lucky enough to get featured again.
We are now developing a free demo with the first 40 levels completely free. Players will get a taste and hopefully be enticed to purchase the rest of the game (and the next 95 unique levels).
It’s an incredibly hard market to release a premium game into but we are quite proud of how it turned out so far. Ellipsis remains our primary focus and we will keep improving it and testing new marketing approaches. If you are interested in the future of this project, follow our blog or our Facebook page, we will keep sharing interesting facts and knowledge gained from our experiences with Ellipsis.