Hi all, welcome back to our developer blog! The video games industry has a tradition of writing and publishing post mortems once a game project has shipped. These are pieces where the team looks back at the production of a game. They help us see what we learned along the way, what we did good and what we can learn to become better game developers. Below you’ll find the post-mortem for Midnight Hub’s first game “Lake Ridden“. Lake Ridden was released on Steam/GOG/Humble on May the 10th in 2018. The game is a first-person puzzler, filled with story and beautiful nature. It takes roughly 7-9 hours to play, depending on how much of the extra story the player wants to explore. The game sits on Steam’s extremely high rating; “Very Positive” and has a 90% positive score, which is absolutely awesome! Midnight Hub is a Swedish indie studio, founded in December of 2015, by Johan Bernhardsson (formerly at Minecraft), Erik Nilsson (formerly at Massive) and Sara Casén (formerly at Paradox). Lake Ridden is all in all made by two game artists, two programmers/designers, one producer, and two musicians. It took us roughly two years to make, at the same time setting up a new studio from scratch. The game itself is made with Unity.
This post-mortem is compiled by the team in August of 2018, and put together by me (Sara – the producer). It’s important to know that much of what is written in this piece is from the perspective of a producer. We’ve also done all our own marketing, community management and sales (with the help of two PR firms around launch). If you’re looking for technical specifications from the development please feel free to reach out to our art director Erik Nilsson or our lead coder Johan Bernhardsson.
Running an indie studio and making your first game as a team is two different challenges, but since they very often overlap we’ll sometimes mention things in this post that addresses both the development of Lake Ridden but also touches on the challenges of creating and operating a games studio in 2018. Our goal with publishing this text is to contribute to the larger body of knowledge about games development out there, as well as take a moment ourselves to reflect on the enormous accomplishment that is creating something from nothing. At the end of this post, I’ll talk about the fact that our game, despite all love it has gotten from its fantastic players and the meticulous marketing effort we did, still four months post-release hasn’t sold close to what we anticipated. This will be a long read, so buckle up!
The development of Lake Ridden started in early 2016, at the same time we founded the studio. To minimize risk and increase success we co-founders created the studio around the skillset we have. We did not start the studio to fulfill one game idea, we quit our day jobs because we want to build a games studio. The game idea that lead to Lake Ridden was one of several pitches presented and the agreed upon by the team. Back in 2016, it was a small scoped horror game, where the player stepped into the shoes of young Marie, searching for her sister inside an abandoned house. This concept would change a lot as development went forward and we got feedback from other games developers on the state of the horror genre. Originally we wanted to make a horror game without gore and blood, with a lot of puzzles and story. But after visiting GDC17 it became very clear that many horror fans wanted gore in their games, and the people who really liked puzzles absolutely hated horror and gore. So one year after production began we decided to pivot the game and went full on with the mystery feeling, the puzzles, and the narrative, targeting the same audience that likes Myst, Ether One and (a much less bloody) The Vanishing of Ethan Carter.
Our team grew from three founders to recruiting an additional amazing junior 3D artist (Anton Sander) and an excellent junior coder/designer (Malin Sandgren). We also signed on a duo of talented musicians; Solid Sounds. Along the way, we secured a big investment, nominations in business awards and front page coverage on big sites like Polygon. Got admitted to one of Europe’s best business incubator. We exhibited demos of Lake Ridden at EGX and Nordic Game Conference, where we had lines of people waiting to play. The game’s visual style got a lot of attention and we had tweets that went viral. Let’s start things off by having a look on what we think went really well in creating Lake Ridden!
What Went Well
1. Excellent User Ratings. The game had an amazing 96% positive rating one month past release and still has the excellent Steam rating “Very Positive”. The game resonates really, really well with its target audience. Making a game is insanely hard. To push a game out into the wild is an amazing achievement alone, doing so and scoring such an awesome reception is absolutely fantastic! Even harder when it’s the first game by a new team. What makes this even more remarkable is that we chose to pivot the game from horror to puzzle-mystery one year into development. It’s extremely rare that a game that changes genre ever gets released. Of course, the game is not perfect by any means, but people rarely believe it’s made by just four developers in two years.
2. High Level of Craftsmanship. The game has very solid puzzle design that challenges the player to really think outside of the box. Reading the user reviews on Steam makes it very clear that a lot of players enjoy the puzzles, the music, the art, and the story. We implemented a lot of interesting design choices in the game like a sophisticated hint-system and a chapter system for saving. The art style of the game has proven to be highly marketable and each area of the game really has it’s own distinct feel to it. The voice actors did an amazing job, as well as the composers. Of course, there are some Steam reviews that rightfully point out things we could have done better (always happy to get feedback!). But all in all, it’s an extremely solid first game for a studio.
3. Creating Inclusive Studio Culture. This one is not only tied to the project of Lake Ridden but since the development of our first game naturally shaped how we worked together, how we solved problems, had discussions and interacted on a daily basis in the team, we choose to list it in this post-mortem. We worked very hard on creating a studio culture that should feel inclusive and respectful for all. We want to build an alternative to crunch, elitism and office politics. Here are some examples of what we accomplished:
- An environment where it’s OK to be wrong. If you have good intentions anything can be talked about and solved.
- Everyone stated their levels of stress each week, so we could keep track of both the individuals and the team’s stress levels. When any team is put under pressure it has a way of ripping the team apart, but we managed to grow even closer during hectic periods since we could lean on each other for help and support.
- 35h work week. We measured the SCRUM points completed and was able to get the same amount of things done in 35h instead of the usual 40h. All while the stress number was staying the same or dropping. Everybody was actively encouraged to use this extra free time to see family or do exercise.
- We deliberately hired people who were very different but shared core values. We wanted to have many different perspectives represented in the team, to better understand our players and complement and challenge each other. This tremendously helped us solve problems and see different perspectives.
- We played and analyzed games together each Friday to help give the team a common vocabulary for speaking about games and game design. This greatly improved the understanding between disciplines and gave us the same references to use when talking about games (i.e “Let’s do the shadows the same way we saw they did in Alien: Isolation”).
- Feedback was encouraged and expected. We put great focus on the separation between whatever feature was critiqued and the person that made it. If you can separate the thing from its maker it’s much easier for everyone to learn and move the game forward.
4. Game Testing With Real Life Players. We spent a great deal of time and money on testing the game on real people. We did this in two ways; on players at conferences and players that came into the office to play Lake Ridden. We closely watched people while exhibiting it on trade shows like EGX, asking them about things like what they felt about the suggested price (19,99€) to what games they usually played. That was complemented with user testing where we invited around 40 puzzle gamers (most of which we didn’t know personally) to the studio.
There we watched them play and asked them follow-up questions from a prepared protocol. We always had predefined hypotheses that we worked with during these game tests, like how long the ideal playtime for a level was, how many tries a player should make before solving a problem etc. If something deviated too much from the desired outcomes we compiled this feedback in an actionable way to the development team. This testing was absolutely invaluable to make sure Lake Ridden got such a good reception from puzzle gamers. In the late stages of the development, we also worked with the fantastic QA company Testology to find and squash traditional bugs as well.
5. Rapid Reaction To Change. As mentioned before we pivoted the development after one year. One of our core ideas of Midnight Hub has been to be able to react to change if needed since few industries are so unpredictable and volatile as the games industry. Changing from a horror game to a puzzle game was a very difficult move to pull off none the less. Pivoting is a very scary and stressful thing to do, and most game projects or teams do not survive such a move. It forced us to do major reworks to the story, the game design, and the music. We still consider it a successful move since it was necessary and we survived to release a 96% positive game after making the switch.
6. Marketing, PR and Open Development. When founding Midnight Hub in late 2015 we knew that the games market was getting more and more crowded with each month. Between 2004-2015 a total of 7 000 games had been released on Steam, with almost 3 000 alone in 2015! So we knew we needed to work with marketing and visibility from day #1 to even have a chance to break through the noise. The first way we tackled this was to have one co-founder with a background in community management and marketing (me). We then formulated a marketing strategy to follow the development to maximize the chances that Lake Ridden would land a top spot on the Steam front page when released (our best bet to set off a snowball of sales). Here are just some of the marketing efforts we pushed along our two-year development (on an extremely frugal indie budget).
– Visited and exhibited the game at EGX17 Birmingham, GDC17, NGC17, EGX18 London and GDC18. We hired booths and took advantage of the fact that shows like EGX have PR firms contracted to help indies market their presence to journalist.
– Johan’s Twitter has 360 000 followers after his time at Minecraft. He continuously tweeted about the development.
– We contacted the press ourselves and managed to secure coverage on sites like Polygon, PC Gamer and tons of minor game news sites.
– We hired a professional trailer company to make a great story trailer for the game which was featured front page on IGN.
– Regular high-quality GIFs, one of which went viral with over 200 000 views.
– We worked with Facebook marketing to spread the word about the game prior to big shows where we would exhibit the game.
– Contracting two really good PR firms to help us secure reviews and coverage around release, both in USA and UK.
– Paying for influencers like Yogscast to play the game on their channel.
– Reaching out to Steam Curators and advertising on Keymailer.
– Competitions at conferences where people could sign up on our email list to win a free game.
– Open development where we often tweeted, facebooked and blogged about the development. We wrote several well-shared pieces about project management, portfolio building and studio culture that mentioned our game.
7. The Launch of Lake Ridden. If you have followed the news you know Lake Ridden has not (yet) sold enough copies to support the studio Midnight Hub. We had to let everyone go in August of 2018. So you might be wondering how we could list the launch as something we’re proud of? Let’s look at what we managed to achieve during the launch of the game:
– Lake Ridden launched on May the 10th 2018. It rose the top of Steam’s front page chart “New & Trending”, and stayed there for five days, in countries like USA, UK, Germany, Sweden, Australia, Finland, and Norway.
– The game secured major organic non-sponsored coverage when both Lirik and Forzen streamed an hour each on Twitch in front of an audience of almost 35 000 viewers.
– Lake Ridden was covered and reviewed by a lot of sites, included Gamereactor and Fandom. The overall reception was positive.
– We got over 2 500 requests for keys on Keymailer and ended up sending out 1 200 keys to credited streamers.
– Big name Twitter accounts like much of the original Mojang gang with over 500 000 followers and Rami Ismail tweeting about the game.
– Front page feature of Humble Bundle and GOG upon release.
– Release trailer coverage on Polygon and many other news sites.
We managed to do all of this without a publisher pushing our game. The only kind of coverage we did not manage to secure was release-day reviews on sites like PC Gamer, IGN etc and we were never shown at E3. After the release, we quickly fixed bugs, implemented controller support, monitored social media for mentions of the game, answered all questions on our Steam page and handed out guides on how to tag the game with proper tags to players that enjoyed the game. The release in itself was something we’re extremely proud of, both in terms of coverage and the technical aspect (extremely few crashes ever reported).
Overall, Lake Ridden is a great game, it had an absolutely fantastic launch and the team learned invaluable lessons while making it. Let’s have a look at what we struggled with, what we could have done differently. Please keep in mind it’s always easy to look back in hindsight.
What We Could Have Done Differently
1. Reality Testing the Concept of The Game Earlier. We should absolutely have tested the core concept for horror + puzzle earlier to get the feedback that it did not appeal to a core audience. The game idea did make sense for us as a team to work on, it appealed to our strengths and a team, and we believed there was an opening on the current market for this kind of game. But looking back we should have tested this idea on real-life players much faster. That could have saved us a lot of work and we could have pivoted earlier.
A suggestion would be to do a paper design of your game or super quick prototypes that you upload to sites like Itchio to gauge interest. At the same GDC17 where we showed of the game for the first time ever, we managed to secure front page placement at Polygon with a 10 minute long gameplay video. At this time we could not pass on such a big opportunity to start building awareness around the game, but chances are this video complicated the game’s development in the end. For the longest of time, this video was the first one people saw when they googled for Lake Ridden, leading people to believe the game was still a horror game after the pivot. An early gameplay trailer or coverage will cement a lot of the game’s sound, art, and design and make it hard to break free from that or change direction or impressions later on. On the other hand coverage like that is super rare to get as a small indie studio.
2. Sound and Music Talent Was Brought In Too Early. We wanted to learn from previous experiences and not treat music as the last thing you smack onto a games project. We hired an amazing duo very early to help us with sound, music, and voice acting. This in itself was great but became a challenge when we needed to change the direction after one year of development. It meant that a lot of the creepy music and mood that had already been set had to be reworked to fit a mystery-puzzle game instead. The final soundtrack of Lake Ridden has gotten a lot of praise among players, but it would have been better for the project if we had brought in sound & music later on when the vision for the game was much clearer.
3. Trailer Driven Development. As mentioned previously we got an amazing opportunity to show gameplay of Lake Ridden on the front page of Polygon in 2017. After showing much of the game in that video we headed into creating the next area of the game. At the same time, we also contracted a really talented trailer studio to make a story trailer together with us. We knew we needed to have a new trailer ready by September 2017 when the game was to be exhibited at EGX Birmingham. The trailer studio was excellent, but since we did not have anything totally unseen from the game to show in the story trailer (and not enough resources to build a scene specifically for a trailer to later just throw away) we found ourselves in a pinch. So the solution became to shoot the upcoming Lake Ridden story trailer in the same area that we had started building right after the Polygon video went live. This lead to a situation where the layout of the new level was largely affected by what we wanted to show in the trailer. This was not the best choice since a lot of that level’s layout ended up driven by the development of the trailer. The layout of a level should always be driven by game design (or storytelling).
4. Huge Scope for an Indie Game by Four Developers. Lake Ridden is a huge game to make on just four developers. It’s an open world game with almost 35 puzzles, 7-9 hours of gameplay. The final art style that was chosen rivals even AAA games in fidelity and detail. We really wanted to make an ambitious project, something that would stand out. Making a game of this size requires both extremely skilled developers and good stress management.
5. Eliminated Effort = Shifted Effort. It’s extremely important to be aware of the fact that if you decide to cut something from the game it might not end up as reduced labor in the end. At the beginning of the game, we had a cutscene planned to show the story of how Marie and Sofia traveled to the campsite. In the end, we had too much for the artists to do, it was impossible to schedule this cutscene into their backlog. So instead we decided to cut this scene and replace it with text made by our story writers. This ended up placing a new burden on the writers instead. It’s important to be mindful of the fact that if one discipline of developments decides to cut something it will most likely affect someone else along the line instead.
We’ll take this opportunity to talk a little bit about the fact that Lake Ridden, despite stellar user reviews and intensive marketing since day one, still hasn’t sold the anticipated number of copies we had hoped. The general advice seems to be careful before you judge a game on Steam that hasn’t been out for at least 12 months, but we’re very interested in why the game hasn’t found a bigger audience in four months.
Much of the following will be speculation and theories, there is no way for us to actually test these hypotheses. We have always believed in being as open as we can about our development, to add to the body of knowledge out there. While we’re unable to specify exactly how many copies we have sold, we can state that it’s less than 3 000 after launching on Steam, GOG, and Humble four months ago. When making our sales projections we looked at games such as Ether One, Firewatch, Kholat, and similar much less known puzzle games. Don’t get us wrong, we did not expect to sell as much as Firewatch which has a publisher and was made by a team five time the size of ours 😉
Keep in mind that we worked with marketing and visibility since day #1. Our marketing efforts include (as previously touched upon):
- One of our developers having almost 400 000 followers on Twitter and tweeting about the development continuously.
- Placing Steam front page in UK, USA, Nordics, Germany etc for five days upon release.
- Contracting two great PR firms at release that secured big coverage among dozens of middle and minor sized press outlets.
- Having Twitch celebrities like Forzen and Lirik stream 1 hour each from the game to almost 50 000 viewers.
- Working with influences at Yogscast to make a really great sponsored video.
- Exhibiting on four big trade shows, having lines of people waiting to play.
- Getting over 2 500 key requests on Keymailer, handing out 1 200 keys to verified streamers in the end.
- Giving over 40-50 interviews for gaming sites during development.
- Four trailers released; story trailer, gameplay trailer, release date trailer, and launch trailer.
- Landing big coverage during development on sites like Polygon, Red Bull Gaming, IGN etc.
- Showing the game off-site at GDC San Fran two times around, for journalists from sites like RPS and PC Gamer.
- Blogging about the development for 2 years, having posts that went viral about game development.
- Having tweets that regulatory got 50-100 likes, one that got 200 000 views and 500 likes.
- Working with Wishlist on Steam for 9 months, always having a call to action (“Wishlist Lake Ridden”) on all our activities.
- Building an email list for more than 2 years.
- Working with Steam Curators to promote the game to groups with 5 000 – 10 000 members.
- Personally handing out 2 000 token cards to real life players that praised the game on conferences.
Theories On Sales Impact
Here’s a brief overview of a much deeper internal analysis on what could have impacted the sales numbers. Please keep in mind these are pure speculations, some more likely than others.
Release in May? We released Lake Ridden May the 10th, with a 10% release discount. We had made sure that no other similar game we knew of would release the same week. The May of 2018 proved to be one of the hottest Mays on record here in Europe, and people don’t tend to spend time playing games inside when it’s super hot. We talked to a lot of fellow indies and they all said that May 2018 had been a really bad month in sales for many of them.
May is also a month when people usually have a lot to do at work or school, before vacations set in. May is also close to the yearly expected Steam Summer Sale. This could have lead to a lot of potential players holding on to their wallets waiting for the upcoming Summer Sale (21st of June – 5th of July, dates not official during our launch). We submitted Lake Ridden to the Summer Sale on a 30% discount, but it did not really raise the numbers of copies moved.
The Price? Lake Ridden is offered at 19,99€. This is a price point we pitched again and again on the real-life players who played our game at big shows. They all said they felt it was a fair price for a game of this quality. This price point puts us in the same category as The Vanishing of Ethan Carter, Firewatch and Edith Finch. Only three of our 50 Steam reviews have something negative to say about the price.
Unattractive Steam Page? We crafted our Steam page based on pages from popular puzzle games like Vanishing of Ethan Carter and Bastion. We tested the content and layout on real-life players and got very positive reactions.
Competition? Maybe Lake Ridden does not stand out enough in 2018? If a player googles “Best Puzzle Games” they will get recommendations for games that have been out way longer than Lake Ridden has. This means that we indirectly compete with classics such as Myst, Witness, Amnesia, and others. And a customer might decide to go for a trusted and praised classic rather than a fairly new indie game even if Lake Ridden has a 90% Positive.
Reaching Target Audience? Puzzle fans are extremely hard to target in marketing. Some of our biggest fans actually messaged us to tell us that they actively unfollowed us the last months leading up to release to avoid getting spoiled on the story. We always did our best not to spoil the story or the solution to Lake Ridden’s puzzles in all our marketing material. Puzzle games usually spread by word of mouth, so chances are that sales will pick up now when people are back to work and school, meeting their friends again.
Technical Issues? We monitored the forums, social media and the Google Analytics tracking ID we implemented on our Steam page. Could it have been some kind of technical issue preventing people from buying? But no one stated they had any trouble performing the purchase in itself. Another aspect of “technical issues” could be that the game is fairly heavy to run, but it’s not extreme by any means in 2018, and nothing we have seen to be mentioned too often by players.
Publisher Push Needed? As discussed earlier the only kind of coverage we did not manage to secure was a spot at E3 or launch reviews on the big giants like IGN, PC Gamer, Polygon or Kotaku. We honestly have no idea if this is the deal breaker for selling games in 2018. Maybe a publisher could have pushed the sales of Lake Ridden better than we did? This is pure speculation and I’m not sure we’ll ever get to know. Some of the publishers we spoke with during development jokingly asked us “what do you need us for when you have managed to get all this marketing on your own?”.
Games Genres in 2018? Should we have added Battle Royal or blockchain to Lake Ridden? Jokes aside, is there still a viable market for single player puzzle games with a deep narrative in 2018? We like to believe that the tremendous success of games like Firewatch and Gone Home would prove so. But maybe players are growing tired of these kinds of games? A lot of what’s big right now are fast-paced action stuff you play with your friends.
Confusing Message? The game started out as a horror game. If you google “Lake Ridden” it’s possible to feel confused about what kind of game it is. This could potentially repel the big group of people that love puzzles but hate getting scared. Then again, one only has to read one or two of the user reviews on Steam to understand it’s absolutely not a horror game. The game’s description even includes the line stating it’s not a horror game nor a walking simulator (which has been a very appreciated move!).
Wrong Publishing Platforms? Lake Ridden is a PC game and available on Humble, Steam and GOG. Would console have been a better fit for this kind of game? Again, this is just something we can only speculate in. We signed with one big console manufacturer quite early on, but since we were only two coders we could not work on a simultaneous release for the game. We would absolutely love to release the game on consoles one day or make it available on Mac.
Indie Games on Steam 2018? Last but not least. To us, it’s baffling that a game with 90% Positive, that spent 5 days on Steam’s front page and managed to grab this much attention hasn’t (yet) moved more copies than this four months after release. It’s a game made by well-experienced developers and loved by those who played it. If we take a look at the statics for Steam that Valve shared on Nordic Game Conference 2018 they stated that the average Steam user buys 8 games/year. This number has been constant for two years now. Between 2004-2016 almost 11 500 games were released on Steam in total. In 2017 around 7 700 new games were released. In 2018 it’s projected 11 000 games will release. Roughly 200 new games are released on Steam each week by now.
The average player on Steam buys 8 games/year and that (probably) on a big discount like 70-80%. Those 8 also include big AAA games like GTA, latest Tomb Raider, Rocket League, Battlefield and more. This means the competition to be one of these 8 games a year someone chooses to spend money on is absolutely fierce. While it’s fundamentally a good thing that more gamers than ever are able to find their niche, without better systems for sorting and recommending relevant games it’s hard to reach your target audience.
A Way Forward
There are many things we’re proud of when looking back at the development of Lake Ridden! Making this game would not be possible without the immense support of the game dev community, our families, friends and everyone who believed enough in us to give this a chance. Releasing a game with 90% positive is a tremendous achievement, even more so when you consider this has been the studio’s first game as a team! As we mentioned on our social media and developer blog we had to make the responsible and extremely tough decision to let everyone go (even us founders need to find new jobs). The praise Lake Ridden got, unfortunately, hasn’t translated into the company’s bank account (yet). Running a games studio of five people is not cheap. Sometimes you simply do your best and give it all you got, you make something awesome, but perhaps the timing wasn’t right. This is a very tough pill to swallow, after pouring blood, sweat and money into something for three years. However, the company still exists and the game is available on Steam, Humble and GOG.
We founders will be showing the game at EGX Birmingham next week, so if you want to enjoy a beautiful story-driven puzzle game make sure to drop by and give it a go! Once again, we’re tremendously thankful for all of you who supported us this far <3
If you have any thoughts or questions on this text, don’t hesitate to comment, reach out to me on Twitter or drop me an email to firstname.lastname@example.org!
Sara -Producer of Lake Ridden