Ahoy, mateys! In this blog post we reveal a lot of background information for those interested in behind the scenes on how our games are made. We believe it to be important to reflect critically on our work and as a part of this process we wrote a post mortem for Planet Buster to summarise our learnings. It was originally intended for internal use but we are now releasing it publicly in the hope that someone might learn a thing or two. It has been written by me (Michael) but it summarises learnings from all of the team.
I will in this post mortem present you with all the inside data related to the development and launch of our casual puzzle title Planet Buster, www.planetbustergame.com.
We have shipped the game on a number of platforms, namely our website(PC, Mac) mobile phones (iOS, Android) and 3rd party platforms(Chrome Web Store, Indievania, Gamersgate) so we are able to compare the numbers between these. Along the way we have also experimented with a variety of sales models such as “Pay What You Want”, price drops and free weekends. You will hear all the juicy details on how well these actually work.
We had three goals with Planet Buster:
- Give us experience with emerging platforms and business models
- Provide funding for our big Pirates of New Horizons
- Act as a marketing tool for Pirates of New Horizons
A lot of reviewers and players ended up really enjoying the final game. The financial side of things turned out less great though.
What Went Right
1. Switching From Java to Unity
Development actually started a long time ago before Exit Strategy Entertainment was founded, at a time (early 2008 to be precise) where the name “PopCap” meant something to a few geeks but nothing to most other people. Casual games just started to become popular and to be a viable business. Our designer/artist Soenke ‘Warby’ Seidel also saw some kind of potential in these games.
He came up with a plan to make something that combined the playing field of “Bejeweled”, that feeling of reward from “Rocket Mania Delux” with the strategic depth of another game him and some friends played competitively at the time: “Puyo Pop Fever”. Soenke was convinced that a such game could be done “over the weekend”…
Planet Buster ended up being developed on and off and neglected for a long period of time. First in a Java version that actually got to be a fairly finished state. It never ran well though (insert all the Java jokes here) and didn’t provide the easy cross platform capabilities for mobiles that Unity does. So that version never really got any attention from anyone nor were there any viable places to distribute it.
This version was then resting on the shelf for a while until Soenke showed the game to the rest of us at Exit Strategy and we decided to port the game to Unity. This port allowed us to have a demo of the game playable directly on our website and also with little effort port it to mobile devices. The Unity port really was what made it possible for this game to actually have some kind of chance of reaching players. The Unity porting job was first outsourced to a contractor.
2. Salvaging a Train Wreck
We did however in the end manage to get the game through to a polished state which I guess goes to tell that it is possible to salvage even a complete train wreck. All along the way it was about striking an important balance though: Had we focused entirely on cleaning up the code to usual standards then we could probably have spent many more months doing just that without the actual game experience moving anywhere. Neglecting code quality and just adding features on top is an approach which I am certain would not have been possible either, as it just is an impossible task to build a solid house on a canoe.
Seeing that we actually finished something is a really good feeling and in particular considering what a mess it started with. At the time of release we had been working on our other big project, Pirates of New Horizons, for about 9 months and still with a long way to go. So Planet Buster was a nice change from that, as the overly long productions was one of my original reasons for leaving triple-A and starting an indie studio. Planet Buster helped to prove the mental idea that even though there might seem to be an endless amount of work left on a game, it is entirely possible to bring it from this state to a polished state with a month or two.
3. Pay What You Want
During the first month where the “Pay What You Want” deal ran we had a total of 153 downloads of the game. This was way less than we expected. We went with the true “Pay What You Want” model which meant literally any price as low as 0$. We were pleasantly surprised by the amount of people who chose to pay for the game though, that being 45% of all people. Also, the average price ended up being 5.9 euro, higher than even our recommended price of 5 euro! After the PWYW deal ended and we had set a price, it didn’t take too long until sales flat-lined. This means that the PWYW model can be a good method to rally your fans, in our case those who would like to support the development of the 3D Pirates of New Horizons game, but as a promotion tool to spread the word it has certainly lost its novelty.
In total a little more than 400€ was generated through our website. Both the amount of people choosing to pay and the average price was very similar across PC and Mac platforms. This is slightly different compared to the Humble Bundles where the average Mac user tends to pay around 20-30% more than Windows users. I wouldn’t put too much into our case though due to the inherent statistical level of error with such a relatively small sample.
In terms of sales the average amount spent on the game is higher than what we expected, so in that sense the “Pay What You Want” deal paid off, but we expected many more hits and purchases.Overall generating awareness for a match-3 title and attracting new users who have never heard of PONH is definitely not the easiest thing to do, but we already knew this.
4. Trying out something new
Looking at just revenue this game is obviously a flop with just a few hundreds of euro in revenue which does not at all match the time spent on it. We did however also have two other goals (platform experience and PoNH attention) and in regards to these I believe we fared somewhat better. Learning at least is an easy one to tick off as we had never published on platforms like iOS or Chrome Web Store before while we now at least know the processes surrounding those. This is an important thing to us at Exit Strategy Entertainment, if we see a platform that looks interesting then we want to give it a try instead of just sticking to the safe area of hardcore titles for PC and consoles that we are lot more familiar with. The experience that we gained can now be applied when bringing our main title to the Chrome Web Store for example.
We have also had multiple inquiries from a totally unexpected front: App stores and OEM distributors in Asia. Without doing anything to spawn this we received several publishing offers for the game in countries like China and South Korea. Did you know that there are three local Android app markets in South Korea alone? We certainly didn’t.
5. Critical Reception
Our mobile release gained more attention which was also expected as that is also our internally preferred platform to play the game. That was not the case right after our first attempt of contacting app reviewers. When we started sending up personalised mails with a full copy to each reviewer we started getting a lot more attention though. These editors receive so many requests, they will skip your mail without losing a word if they can’t try the app immediately. Save yourself the trouble and include a Promo Code or download link to your Android app right from the start.
Reception was overall very positive with reviews such as “Planet Buster easily holds its own against the well-known Bejeweled and many may argue that it has surpassed that game with the subtle changes and challenge it has brought to the table.” (Indie Game Magazine).
Also getting thousands of people seeing the Pirates of New Horizons name has to mean something in terms of brand recall, we are however unsure of exactly how much this means or how to track it precisely.
What Went Wrong
1. No Game Gets Made Over a Weekend
When it was decided to port the game to Unity, most of the dev team was busy building a company around our 3D platformer Pirates of New horizons and it was therefore decided to outsource the Unity port of Planet Buster. Also, the later was just supposed to be a quick and small side project and not take any focus. Internally in the company we had just Soenke (designer/artist) associated with the project and him together with the contractor worked on and off on the game for a while.
Now 3 or 4 years after the development started we can tell that no good game gets made “over the weekend”. Also while it started out being quite different then with each revision it moved closer and closer to the always popular Bejeweled. So close that at a first glance it may look like a bejeweled clone, which offends me a little when people say it (as there still is quite a number of unique twists on the concept), but I guess we should have stuck a bit closer to the original vision rather than listening to outside feedback which desired the design to be closer to Bejeweled or match-3 genre conventions. I guess you’re damned if you do and you’re damned if you don’t.
If you think about developing and releasing a second project be aware of what you’re getting yourself into. Tinkering on a pet project is one thing. Releasing a balanced, bug-free and entertaining game that you can sell to players with good conscience is a different one. The game also needs to be marketed and will be a distraction, which can be very dangerous for small indie development teams where resources are spread thin already. No matter how small the scope is, the game will be a distraction and every minute that you spend on it is a minute lost for your main project.
2. Outsourcing All The Code
When taking the project in-house we learned that the code base was a huge mess and I spent quite some time fixing it up. One major code issue was that so much was being compressed into one place, e.g. one JS file spanning three thousand lines of code. A lot of the comments in the code and some of the variables were not even in English. I used to be university lecturer so I’ve seen a lot of ‘funky’ code in my time but this code base was really in a bad shape.
There was a time where bugs and hacks seemed to be surfacing at a faster rate than I was fixing them, ie. fix one piece of code only to realize that this piece of code was broken in order to patch an underlying component which was flawed. The code base had clearly been developed with what I like to call “Photoshop coding” where every bad brush stroke is covered up by adding an additional layer on top.
If we ever outsource any programming again, we will have much greater focus on:
- Performing an in-depth analysis of a contractor and
- Tight tracking of the code base and no payouts before the product AND code base has been verified as reasonable
Outsourcing anything without an in-house expert tracking that component is definitely a no-go.
3. PC/Mac Sales
The Pay What You Want offer didn’t reach as many players as we had hoped for. We tried the services of an agency specialized on spreading the word about indie games for PR. They were new and had a launch offer to promote the games of the first 10 developers who signed up with them. Unfortunately their press release resulted in only one review.
A second round of e-mails sent by us including direct links to the game turned out to be a little more successful, but in general it has proven to be rather difficult to generate awareness for a match-3 title despite new mechanics and great artwork. At least as long as your marketing budget only consists of a Google Ads coupon and an exclusive reliance on PR and word-of-mouth.
4. 3rd party PC platforms
We were a part of the alpha program for the Unity integration of Google’s Native Client which allowed us to bring the game to the Chrome Web Store. The team at Google was really great to work with and they were always quick with thoughtful responses both on the sales and engineering side.
Making the NaCl port work took an awful lot of time though. We were working with an alpha version of the Unity NaCl component in an early beta of Unity to deliver a build to be executed on the development versions of native client running in the beta of Chrome. Thankfully most of this work were to be done anyway for Pirates of New Horizons so Planet Buster was mostly used as a testbed due to a its smaller project size and at this point stable code base.
We sold a total of 31 copies there, spread out very evenly over three months. The lackluster sale here is probably heavily affected by the fact that we at first did not have a free demo version directly on the store, so if people wanted to try before buying then they had to do this at our website. Going with in-app purchases would obviously have been the right thing to do but at least a free demo is an absolute minimum. We did a price cut after a few months but this did not seem to affect the rate of sales on this platform.
The handful of units sold on Indievania, Indiecity and GamersGate are hardly worth mentioning. Same as with CWS this was not worth the effort – at least not in terms of immediate revenue which in total for 3rd party platforms was less than 100€.
5. Mobile Release
The game was released for PC/Mac via our website in late October. While using Unity allows for easy cross platform development which had allowed us to play on our iPad from early development then it still took some work to get our mobile release ready. iOS was not so bad but for Android we had to fix a number of platform specific bugs. Due to us also being occupied with work on Pirates of New Horizons this ended up delaying the the mobile release until December.
One important App Store revelation with regard to release dates: the day that the app is approved is the release date! The AppStore doesn’t care about when you flip the switch to release it to the public. We waited some time until we set the app free (waiting for us to get rid of the last Android bugs), the consequence: no chance to be in the “new & noteworthy” category, because the “release date” goes several weeks back. A google search revealed that we were not the first ones to do this mistake. I mention this here, in the hope that other indie developers won’t follow the same fate.
When it came to publishing it on the iOS App Store we also stumbled into an issue: None of us are Mac users and while Apple devices might be relatively easy to use then a lot of the dev tools are definitely not. It took a surprising amount of work to get all the practical details regarding Xcode, provisioning profiles, certificates, app submission, etc. working properly.
We did a free weekend which made more than 25.000 people download the game. Quite a lot more than we had expected but it appears that there really are lots of people who visit those sites that track “apps which went from paid to free”. A ‘free weekend’ on iOS means that it is the user’s to keep and play forever in contrast to e.g. Steam where access is shut off after the weekend for those who did not actually pay money for it. We were therefore counting on the word of mouth and general attention around the app to bring in some sales on the Monday following the free weekend. On that day however, we sold zero copies. We did however get some bad reviews out of it from disgruntled customers who had paid for the app just a few days before.
Experimenting with different prices is easy to do and we highly encourage this until you find the optimal price point for your app. Free weekends on the other hand didn’t have any notable positive effect on our sales and you run into danger of disgruntling customers who paid for the app not long before.
On Android we only sold a fraction of what we did on iOS (29 vs. 371 copies) – total mobile revenue ended up a little over 400€. This might be related to the fact that we got press coverage from a lot of iOS exclusive sites, a corresponding phenomena which hardly occurs on the Android side.
Reviews stand in contradiction to the revenue as being stellar across the board. This could potentially indicate that while our game on its own is strong then we might need to put more focus on developing a viable business model around it and or decrease the turnaround time on each project and push them from idea to release much faster.
In terms of rallying supporters of PONH, giving them an opportunity to support the game’s development financially and in return give them a fun game to play in the meantime, in that regard I believe we succeeded.
We just recently made it export to Flash and we are now wondering if it would belong on a place like Facebook or Flash Game License. This along with the Asian distributors might provide some kind of future for the game still.
Overall we believe that Planet Buster was simply just too late for the party, both in terms of gameplay and business model. From the time we originally decided to make a match-3 game hundreds if not thousands of similar titles had been released. When we got around to shipping our game most of the revenue had moved to F2P models.
Developer: Exit Strategy Entertainment
Publisher: Exit Strategy Entertainment
Release Date: October 25, 2011 (PC/Mac), December 16, 2011(iOS/Android)
Number of developers: 2 full-time and 1 contractor
Length of development: 2-3 months stretched over 3+ years.
Development Tools: Unity, Photoshop, MonoDevelop