This post was created to help everyone that is about to release a new game in Steam that has never done it before. Also, even though post this is based entirely on my experience with Steam, your experience might vary.
Even thought this post doesn't really talk about development itself I can assure you that it was a very intense experience. I had to build the whole engine, deal with thousands of bugs, gameplay balances and such. All in total almost 12.500 compilations in 17 months, you can do the math. My obsession for perfecting gameplay gave me thousands of attempts to see if the game was more playable after every update.
This document will show you all the mistakes I made during my latest title publishing but also its amazing evolution to a better game.
Let's start the journey!
Avoiding the Beekyr catastrophe again.
In 2014 Kaleidogames (or just I... I'm just one person!), decided to make our second shoot'em up game.
Our first one, Beekyr in 2013, was a pretty hardcore (but a lot of fun) game, adapted for Android with touch screens & gamepads, iOS, OUYA and finally for PC. Also, there is web (flash, lo-res) version I have uploaded here. It had several thousands of downloads but not many sales.
The game didn't sell much due its strong theme dissonance; a hardcore game with kid's orientated insect graphics. That attracted the wrong people and repelled our target players: shmup experts. I went through this behaviour in a previous blog post called: Disappointing stats about Beekyr.
This time we were going to avoid the problems with the theme and target competitive players. It would be based in scoring as much as possible instead of advancing through scenarios. This one would be more like old space shooter games like good old Galaga, Space invaders, etc. Its name was Vortex Attack.
Take a look how the game evolved from September 2014:
PC in Madrid Games Week 2014
To February 2016:
Two dedicated machines for fairs and one of them resident in an active arcade venue.
Presenting Vortex Attack in fairs: first impressions.
After nearly 2 months development, I took the game to a fair in Spain (Madrid Games Week 2014) People liked it, I even had someone asking me to put it on Steam. I felt joy but the game wasn't ready. A couple of months later I took it to two more fairs and more people mentioned Steam again. This time they were more pushy, they wanted to play at home.... I also had some kids playing non-stop in one of the fairs (Parlabytes 2014); they were totally hooked by the game. They literally told me, with red eyes, that they had never played a game for so long non-stop, they loved the game.
To be honest, I felt like a drug dealer, and they wanted to have my game, their dope. I have never experienced that kind of situation from this side but it was quite interesting. When I was a teenager I did the same with Starcraft and Heroes of Might and Magic 3, they were my dope back then.
The kids playing
So I felt the game was already advanced enough and I decided to have a go to Greenlight.
Road to Greenlight
I started preparing the marketing assets:
Logo version 0.7
First Cover art (I liked It but I knew it was too simple)
First game trailer.
Back then, that was all I had to show. I felt that it was an amazing way to show my game. I had the impression my game was really good!
Yep, that was me losing the objective vision of the game! I had been working 10h a day on average for 4 months to get like this. NOT RECOMMENDED
By the way, have you ever wondered why really shitty games with shitty graphics & shitty gameplay get ever released? This is probably the reason.
Also, the addicted players that already played, had no barriers to get into the game. They just had to sit and play with a gigantic x-arcade controller. I have to admit they were a bit reluctant to play it at first: not very attracted to the game at first glance but once they played for a couple of minutes, they didn't want to stop playing. The rest of visitors & players gave good feedback:
This meant something that I didn't figured out until it was a bit late: They were biased after playing the game and talking to me.
POST MORTEM ADVICE HERE:
Our mistake: Thinking that people online would react as positive as people that have played the game already.
Result: I wasn't going to be there with the user viewing the game online, so their opinion would be entirely based in the assets I was going to upload. Players had an early opinion of the game and obviously, they thought the game would end up very similar to that.
How to do it correctly: Ask real (not virtual / online) people to give you feedback on the material that you are going to display online. Don't let them play your game or hear any comments from you or anyone. That feedback will be priceless. But be warned: get ready to be hit with a brick right in your nose, specially if you push them to be 100% honest with you.
Placing the game in GreenLight.
I finally went ahead, paid the fee Greenlight new account (£70/ 100US$), created a new Steam account and placed Vortex Attack in Greenlight. For the first 2-3 days Vortex Attack got about 800 positive votes. It's quite good for a game that hasn't been in the media, just in fairs.
HOW DOES STEAM'S GREENLIGHT WORK?
Steam needs enough of those positive votes to put someone from Valve interest in the game. This is not mathematical, the Greenlight system is just a pure reference for them to consider games entering their market or not. But don't worry, these days, the filter is pretty easy to pass. Back in 2014 some good games could take 6-12 months to get in, my game wasn't that impressive and took a lot less time.
After that, Vortex Attack was moved away from the main page as new titles were coming in, and most of all visibility went away and votes immediately just stopped growing. From that moment it depended how well I told the people about that Greenlight's page in forums, Twitter, Facebook and email.
Luckily for me, a few days after there was someone, Alie from Groupees (a site about bundles) telling me that placing my game in his bundle would help it to get extra visibility. Before of that, I had to upload my game to the Desura (indie portal) in order to be able to sell my game in his bundle. So, I talked to a friend that did the same (@ludipe with Missing Translation) and I finally took the risk: We placed Vortex Attack in the bundle: "BAGB18". I just had to wait a month for the bundle to become active but it was great. We sold almost 4000 units; from that about 25% of the people took time to vote the game in Greenlight.
Note: the alpha of the game had to be downloadable from Desura but when the game made to Steam I would have to give all supporters a Steam Key.
At first, I thought it was pretty good as it ended up increasing my visibility permanently. I didn't have to nag people anymore, I could focus back in the game development. And after a few weeks it happened suddenly: Vortex Attack was greenlighted (2 months time in total).
Here is how it went (it's the green line):
You can clearly see the first few days, then nothing happens till the bundle becomes active.
There is another interesting post where I explained other details about the process until here.
Check it out here: How did we get Vortex Attack rushed into an Early Access version
THOUGHTS AFTER BEING FEATURED IN A BUNDLE LIKE THIS:
Later, I would realize many of those keys were actually being traded (even sold) in some websites. I suspected sales would not rise much for a while until those people run out of steam Keys. It's silly but there is speculation even for a 5USD game. To avoid this kind of thing Steam should let us set an expiry date for the keys. People that get them shouldn't be able to trade them, it 's not fair for small indie studios with very low prices.
Anyway, I was working on the game adding levels and better AI systems, and still adding playable ships. But with the possibility of adding my first game to Steam I felt the urge to release the game ASAP, there were very exciting times.
Greenlight passed: road to early access.
I thought I could get the rest of the game done in about 2 months, after all it had already most of the features and I only wanted to release something playable ASAP.
In order to add the game to Steam, I had to do a lot of work for the market place, this meant time not progressing with the game.
POST MORTEM ADVICE HERE:
To publish a game in Steam you have to do all of these tasks, I didn't know back then but I should have allowed more time to prepare all of this, it was a very stressful time to get all of this done as well as the game... Please allow at least one month (ideally 5+ weeks) to prepare it all.
1 - I prepared the Steam page with all the game information:
2 - I learned how to implement Steam features which by the way aren't necessary for publishing in Steam, but they can add a lot of value to the game.
3 - I had to learn how the SteamSDK's build system worked and how to upload and publish the game, which took me little time. I have to admit once I created a .bat file it's possibly the fastest publishing system out there.
... but also I had to prepare more things to get noticed by the press.
4 - Created a new page for Vortex Attack at www.kaleidogames.com
POST MORTEM ADVICE HERE:
Some developers stop here and that's wrong! If you don't tell the world you have made something amazing they will never notice. This is as important as development. Trust me! The game won't magically spread though the web, I have gone through that already , and I can tell you, it doesn't happen that way. Unless your game is featured in a shop or something like that, but chances are way to low to take that risk.
6 - Go site by site gathering contacts information to who I should send the press kit. This is an ongoing database, you should start this ASAP, it takes ages! But to be honest, I would not do this step If I had more economic resources. If you do have a little bit of spare money you should not get involved in this kind of thing, and should pay someone that can do this for you and send the press kit for you, they probably have a data-base already. You would still have to write it but finding contacts is possibly the most tedious task of all of them.
To accomplish all of this took about 3-4 extra weeks of very intensive work. So I had to I delay the release date a whole month from its initial release date.
But I thought I could show the game as beta before of release date and I ended up uploading it as 'Early access'. This happened before creating the web and sending the press kit ... I thought nobody would notice and Steam wouldn't make a fuss about it. I really thought it would be quiet and perfect for testing the game but this is what happened:
POST MORTEM EARLY ACCESS STATS (27 March to 27 April 2015):
For the first month it had:
- Almost 1 million impressions
- About 25 thousand page visits
- Sold: About 50 units
From which: 70% of those done on the first week.
- Steam keys redeemed: ~1.600 activations (from the bundle)
A couple of weeks after 'early access' release day, there weren't many downloads, it was pretty much stalled. There were a few people involved in testing and balancing the ships, power-ups and enemies. I was also able to find usability issues and game exploits along with a few some other issues: in this sense, 'early access' was working great.
While I was working in all of this, I sent some friendly emails here and there about the game telling about this WIP version (Work In Progress) but some people reviewed the game as if it was final or close to final.
POST MORTEM ADVICE:
Our mistake: Sending emails to some people to just check the game out for hype or personal opinions.
Result: Early judging. To be honest, reviews weren't that bad but they didn't reflect the final version. What is worse: those reviews would remain active forever, and people would find those later even with the whole (final) game released. That would confuse people and others will think is a unpolished game with poor graphics. Luckily in one of them, I had the chance to send a second revision to a retro Magazine (view here) before it was finally published.
How to do it correctly: If you want to generate hype, show some screenshots but show as little as possible, just to awake some curiosity.
I believe is best to show your game in fairs: Let the people play it, you can tell them what is being changed and what is in early states. Most of them will understand what you say, some of them don't understand about game development at all and they will assume changes will be minor but if their experience is good enough they surely will have some interest in the final version. Let them play without your input, analyse how they play, what they don't understand and then, ask for feedback: Figure out what they like or dislike. Also, very important, don't lose them: ask for their email to notify them when the FINAL version is out.
Also, very important, if you are showing your game online, in a magazine or anywhere you can't directly interact with the viewer; graphics should as close to final version as possible. In my case, game graphics were extremely un-detailed and primitive, almost as place-holders. I believe that many people received Vortex attack as a crappy little game. It doesn't matter how fun and smooth it plays, if their first visual impression is bad, it will remain bad for a long time, and that is pretty hard to revert!
So, in this case, lots of people got the wrong impression. Letting the people play the game too early might work well in fairs but not online, especially if they have to pay for what they see in screen, was it worth it? Most people didn't think it was (and of course it wasn't!). So after this first impression people can recognize the logo of the game and will avoid it in the future.
Release version 1.0 day.
Time passed and I improved the game and updated to version 1.0 (in May), some of the upgrades were:
But I also, worked in the engine to improve the experience overall:
* This is the first cover art I made for Steam:
Again it was too simple and didn't say anything about the game.
So decided to have another go:
This wasn't looking good, it was too messy and looked a bit crap.
I needed a real illustrator that could make something better than what I could possibly do:
Fortunately, after a couple of weeks I found an artist to help me out, and he managed to send me the cover a week before release:
Artist's cover based in the primitive ship models.
This cover job happened in April. Unfortunately, right after, we improved most of the enemy ships so the cover was already outdated when version 1.0 was out.
After all of this: I had to create a new video with the new game-play and its graphics, this is the best I came up with. This time, took me a couple of days and I wanted to show more the true spirit of the game craziness:
Note that I have been saying 'version 1.0' but not "final version". There were even more bits I wanted to improve but I convinced myself that I had to release the game back that date. I didn't want to keep going on updating forever like happened with Beekyr. Sadly, it turns out that after that day I released another 4 major updates: July, September, November and January 2016...
That release month (May) was extremely stressful, I had to make sure the game and press kit looked as clean and up to date as possible. Also, I had to re-check everything in the game for a few times, making sure game didn't break, stress it, test it, play it with friends, send it over to testers, wait for their feedback and errors and go to sleep at 4.a.m most of the days.
Also, that same week, I had to travel to talk about the new Vortex Attack Arcade Machine in an festival called "ARCADE CON 1" in an Arcade Room called Arcade Vintage. Then, a week later go to another fair. Of course all the feedback I was receiving was very positive. I felt almost like a rock-star but dead tired.
The arcade machine in Arcade Vintage
But, everything is not all wonderful, dealing with all of this along with my also busy personal life with my son (<1 year old) was really hard. I had to sleep little time and become not very social for weeks to get things done. I had rough times with my partner but in the end it all ended up well.
Going back to the release day in Steam....
In the steam Store, admin page, I clicked the Release button what would un-tag the game as "Early Access", update the steam page with the final contents and place a banner of the game in front of 1.000.000 people.... This is a critical moment in the release of any game in Steam. Especially if you have little budget for marketing!
Want to see how did it go? Here are the statistics:
RELEASE VERSION (1.0) STATS (27 May to 27 June 2015):
For the first month we had:
- A bit over 1 Million impressions (~75% first week)
- About 18,000 real visits (~65% first week)
- Sold 3 times as much as the 'early access' day.
- Steam keys redeemed: Over 200 activations
After these stats, I think you should know more about sending the press kits to the media:
A week before release, I sent the first wave of press notes and after a week a second wave to only the people that didn't read it already. There was quite a lot of opening rate on the emails. It was a near 30% and we even made it (briefly) into IGN Spain. Unfortunately, most reviews were only done by small sites. It appeared in Siliconera but it wasn't a review, just a direct copy from my press kit. Most reviews scored over 7.5/10 in many places except for one Italian site that gave a 6.5/10 because game was far too random (gave me motivation to fix it later). One American video review that liked the game*1 but hated the graphics.
POST MORTEM ADVICE:
Our mistake: Releasing the game, once again, too early and to make it worse, sending the press kit to almost 700 professionals.
Result: Early judging again, this version was more approximate to the real final product but still needed many graphical and gameplay tweaks. As result we had one reviews telling that some some ships (player and bosses) looked like place holder graphics*1 . I must admit there looked very flat and poor, I never got round to improve those graphics as I was super busy with the actual gameplay. I believed it was what it mattered mostly, but realized that graphics were more important for everybody... Anyway, I fixed those graphics later in version 1.1). But once I fixed all of that, the review was already done, and media had no interest in publishing an updated review.
How to do it correctly: I should have waited for another couple of months more and probably uploaded this as the original and first "early access".
I believe now that Early access should be used for uploading betas. And I mean real Betas, with most of the final graphics and features, just testing and game-play to fix. And very important: avoid publishing if it has very primitive aspects of the game like mine had.
After this version, I didn't see a big impact in sales or downloads. There were certainly more downloads than before for a while but not the numbers I had in my head.... Trying to boost downloads, I decided to work in the game for 6 weeks more and add some cool features I had in my head:
Even with this update, sales didn't improve that much either, there wasn't any press kit sent or any extra visibility: this depended entirely on Steam or me posting in blogs or twitter. But a few days later after I was on holiday. I couldn't push for more sales and I didn't go online for a few weeks so sales didn't rise too much. I didn't feel guilty because I really needed those holidays!
After my holidays, in September, the tournaments were starting and I tweeted about them quite a few times. I didn't see any significant extra activity. I got cool sponsors that gave away lots of prizes (click to see them) but I believe not many people knew about the tournaments. Also, the button (inside the game) wasn't labelled. I found out that people are not very curious, most of them didn't click it. Note: To make things clear, I ended up adding button tags in a later update.
I went very active in social networks: I personally spoke to some people about the tournament and they were happy to join (they are free!). In the end, there were only few participants, in fact 90% of the participants were the people I spoke to... I still blame the button not being clear enough and that Kaleidogames not that well known outside the Spanish Retro scene(even though there are about 4000 followers in twitter). I tried to make more noise but it's definitely not enough with so many studios also trying to make noise as well. Sadly, there were some amazing players that never made it to the tournaments. But I'm happy that the few players really played hard to improve their records.
Along with the tournaments I finally ended up improving the logo and cover art with the new ships graphics. I think it was greatly improved from the original!
Note: Original was pretty nice but that hand made feeling didn't look right for this specific game.
Final poster, based in the final in-game graphics (1.1)
Before this last cover update. I wanted to make the final version of ALL graphics, including player ships. So I spent a few days improving all graphics and animations from the game, this includes player ships, bosses, shoots, power-ups, vortices, enemies and particle effects.
You can see most graphic updates that player ships have suffered across all game development from September 2014 to November 2015 and how those abstract shapes became more mature and detailed:
Version from the first Alpha (October 2014)
Version from the second Alpha, animates lines added. (November 2014)
Version from Early Access, shapes and shadows tweaked (March 2015)
Version from version 1.0, some detail added (May 2015)
Final version from version 1.2, detail, ship parts and shadows improved (Sept 2015)
Includes some extra unlockable ships from higher levels.
POST MORTEM ADVICE:
With every graphics revision I was unconsciously convincing myself the graphics were fine. I was telling myself some excuses that made those graphics valid. Now, after seeing these updates I can clearly see they weren't good enough. If you are working in your own this will probably happen to you as well. Remember to be objective as possible, because if you aren't, the game will probably fail.
At this stage I was finally happy with the visuals (and the rest of the game) and I ended up sending another round of press kits (~700 emails) but almost nobody answered/published anything (about 2-3?).
POST MORTEM ADVICE:
Our mistake: Sending a second press kit for the same game.
Result: It doesn't matter how much better the updates are. The media won't appreciate any of this, they want new content not new revisions. They have NO interest in reviewing it again. This would happen if they feel this update would be interesting for the community. IE: An update as a DLC in a popular game might be published but should not happen for a small indie game. Unless they personally know you, they might do in that case.
How to do it correctly: When sending a press kit, make sure that you will be sending a final version. Don't do it if you feel that you can improve the game, we all have limits but it is better to hold for a little longer if you will you can do better in a near future. If you think you will improve it, then I would like to emphasize: Keep the game unpublished or as 'early access' (if you already are) until you can display a fine product with the real final art.
I will say it a different way: keep yourself calm, don't rush to release a game too early. A good way to know if you are in good track is to stop working on the product and work on something else. This requires to not open the game or to see anything related with it and don't view any in game graphics. Then you will be able to regain some objectivity, this is a really hard thing to do but it might necessary when you work alone.When you work in a team, it might be a bit different.
Keep calm and release when it's ready.
Adding Steam's Big Picture to the game.
A friend noticed Vortex Attack wasn't appearing in "Steam's Big picture" because the game and menus had to be fully usable and playable with a game controller. And that was pretty important as many people use Steam this way!
POST MORTEM ADVICE HERE:
Try to implement Steam's big picture ASAP, you will gain visibility.
What is Steam Big Picture? It converts Steam's client to full-screen and your PC will interact as if it were a game console. If users are using this mode in a TV room then controller support has to be total, that's why it's important. Check it out:
This is "Steam - BIG PICTURE". Pretty nice, isn't it?
Back to the fairs - Updating to the final versions
I took the game back again to Madrid Games Week (2015). This time game was polished and almost complete, with just a few tweaks left. We took a custom made bartop and people were instantly in love with the game*, I knew It was close to the final version but I decided to take more notes of any problems that were still in the game. At this stage I thought it wouldn't be any issues but turned out that I could improve some bits specially related to new players. The game requires the player to know something. Instructions were right in front of them in the screen as a massive text banner, but they didn't see them because they were focused on gameplay. I ended up placing those instructions with some attractive images and stopping the game for a couple of seconds: that worked perfectly. This is one of this biggest reasons why going to fairs feels like a massive play test.
*There will be always someone that won't get the game. If that number is low you shouldn't worry. There are some strange people in fairs!
POST MORTEM ADVICE:
Our mistake: Thinking that people read text on screen even though we only showed that text when they did something wrong.
Result: People get bothered by text on screen. No matter how important those are.
How to do it correctly: If you don't want to bother users, try to display the tutorial text ONLY when the player plays incorrectly. But also, place that text in a clean environment and make sure nothing else is happening in gameplay: no bullets, enemies or auto-scroll should be there and stop them controlling the player for a second. Otherwise the hint will appear in the screen, will become a pain and they will still won't read it! The only solution I came up with is to PAUSE the game while you transition the text in, and make it very noticeable. They will know that box has just appeared and then wait for a prudent time for them to read the text without having to rush. I believe this is the only way to display instructions in a frantic shmup game. If you put them in a different way, you might fail. (But, if you do it differently and you are positive it's working please let us know!)
After that fair, a few weeks later, I was presenting the game again in ParlaBytes (2015) , those kids that played my game a year earlier, came with many more friends just to play it in our customized bartop. They wanted to play again... a lot!. Those were Vortex Attack's first fans, and now they were experts of the game, they even made a steam guide with all details about the game.
Me and my first real fan, he's wearing a KaleidoGames t-shirt! Such a good feeling.
NOTE: I pixelated both faces as you can imagine, he is not old enough for me to put a photo from him freely on the internet. It's not fair on him to just disappear from this moment so I disappear too.
They played that last version and I didn't have to improve or fix anything; they had no requests. A couple of months later they went into the game in-real-depth to create a Steam guide and found a few issues that I gladly improved or fixed. Such a threat!
After every fair with the game in Steam, I was able to tell people about the game being already available to download, they bought the game from fairs or from home in Steam. Always providing an alternative, not everyone knows about Steam or want DRM. If so, they can get a copy from itch.io which is DRM free but it has less features. This is specially good for home arcade machines.
People really had a lot of fun in all fairs. I ended up compiling a new video about their experiences:
Video : Vortex Attack in Fairs compilation
Sometimes, if I was going to exhibit Vortex Attack in a fair with the Arcade Vintage crew, we would be displaying the game in two dedicated machines at once: a bartop and a Full Cabinet, arcade machine:
At this stage, I felt the game was complete, no more updates were necessary...
But, as usual, something else happened!
Tournament rules were being exploited and I had to fix them if I wanted to avoid funny/hacky hi-scores and also I tweaked gameplay more (for its thousand time) to fine tune it again. It was good but there is always a way to make it better. I couldn't miss the chance. Version 1.2 was here and this time I was convinced that was going to stay forever.
Also here is the main menu layout evolution:
I felt confident with the game again. I needed a boost of visibility. So I pressed the famous visibility button of steam for a round of visibility.
POST MORTEM ADVICE:
In Steam's control panel there is a button that will place a banner of your game in Steam market somewhere in the screen in front of 500.000 users (or more if you are doing great IE: selling a lot). Normally this will take about 5h to complete and this doesn't mean players will click it or see it but it is a small bump in sales. It is intended to be used after a big update. Don't rely on those buttons too much, sometimes they can be disappointing as you can see:
VISIBILITY BUTTON STATS:
After this, nothing really that special has happened. The tournament ended happily and new players joined after it.
After reading all my stats you figure out how well the game has sold and when. It's interesting how the game is not currently stalled and it doesn't have the typical long tail.
Here is a sales statistics since "Early access" in April 2015 to January 2016:
Steam sales stats
The game still going with new players spending a fair amount of hours in the game. Some people have played for over 120h. Also, I can tell the few people that are able to find the game and buy it are very happy with it. Vortex Attack currently has positive ratings with over 88% of total votes. For me this is insane but it makes me feel confident: KaleidoGames have created another fun game. This time has been a lot more successful. And after this experience, I have learned that I surely not be making those mistakes again in the future.
Finally, in February 2016, I made another round of interesting upgrades including gameplay improvements, new stage types, enemies and a new set of Steam banners (they call them capsules):
With that, I uploaded the latest upgrade, version 1.3, hoping to leave it like that forever. Always hoping some famous guy discovers the game and increases my visibility enough to sell a few thousand copies, but chances are low as winning the lottery.
If you are a reviewer, you should get an honest email from me about the latest version game. If not but you want to give a hand, please email me and I will send you everything you need to cover the game. I would really appreciate it.
To be honest I feel we should always keep our hopes and moral high and keep triyng, otherwise it would be impossible to be an Indie Game Developer...
I hope you enjoyed this document as much as I did writing it.
If you want to take a look or contribute please get the game. It's a shmup / retro arcade shooter game, but I can assure you it's very playable and fun.
I will be pushing this game in the future for a little bit more, it well deserves it even if it doesn't have the support from the media.
I hope you liked this post, it has been a pleasure sharing all of this with you guys, I hope you don't mess it up like I did and appreciate all the information I shared with you.
This is the end the post-mortem. I hope some you have found this information interesting, if you want to contribute with more information or statistics post them underneath, we will appreciate them!
Note: This post was originally called: Postmortem: Vortex Attack but I considered that a postmortem from a game you never heard of wouldn't be interesing for you! You can read the original here.
More info about Jaime or Kaleidogames:
Vortex Attack site: http://www.vortexattack.com
Kaleidogames Facebook: https://www.facebook.com/KaleidoGames-1537985409799803/?fref=ts
Vortex Attack Facebook: https://www.facebook.com/VortexAttackArcade/?fref=ts