Terminus: Zombie Survivors recently emerged from Early Access (EA) to 1.0. Look at this review chart for the game.

When Terminus had its original EA launch in April of 2021, it didn’t look like it would be a hit. The game only had 15 reviews in its first month and it took almost 1 year to earn 100 reviews. That huge spike on the right side of the chart is August 19th, 2024, the moment the game launched to 1.0 after more than 3 years in Early Access. Since 1.0 it has earned 800 reviews in 10 days.

My typical advice for Early Access release is that you should have the mindset that your EA launch, is your launch. Launching to EA is not a “trial” and it is not “testing the waters.” That is because TYPICALLY, your first month of sales is your best month of sales. If you have poor sales at EA launch, you have nowhere to go but down. 

I did a study (with the help of VGInsights) where I scanned every single game that had both done an EA launch and transitioned to 1.0 to see how they performed. I found that games that don’t earn at least 50 reviews by their first month has a 1% chance of later earning 1000 reviews by their 1.0 launch.  

Similarly, the best practice is to get your “coming soon” Steam page live at least 6 months before EA launch so that you can collect at least 7000 wishlists so you can show up in Popular Upcoming.

Terminus: Zombie Survivors didn’t follow conventional wisdom on either of these fronts. Their “coming soon” Steam page was up only for the minimum 2-week period and launched with only 100 wishlists. Their first month after EA launch yielded only 12 reviews. 

By any conventional prediction model, this game was dead on arrival. There was no chance. But somehow, after years of perseverance, Terminus’ 1-person development team has built up a real player base. As of August 30th, 2024 the game has 2,241 reviews and has a concurrent player count of 2,130 and it just keeps going up.

In today’s blog I interviewed In-geon, the developer behind the game, and analyzed the stats to try and understand what factors allowed Terminus: Zombie Survivors to beat the odds of EA and actually succeed after a slow start.

Steamdb chart of the lifetime concurrent player count for Terminus:

What is Terminus: Zombie Survivors 

Terminus is a top-down, pixel-art, turn-based, randomly generated, zombie survival simulator. You must pick a job role (soldier, police officer, firefighter, farmer etc) and then travel from a random spot on the map to the “Terminus” which is the safe zone located at the end of a railroad track.

Navigating to any node on the world map drops players into turn-based environment where you can explore, loot supplies, and fight zombies. 

The game is endlessly replayable and each run is different. 

This is the classic type of game Steam players love: crafty-buildy sandboxy games that are non-linear, endlessly replayable, and infinitely complex. I mean, look at this cooking and crafting menu! You have 3 different types of fish and 3 different types of crustaceans you can use in your recipes. (Side note, I love the idea of traversing the Zombie Apocalypse with a lobster and a Dungeness crab in your backpack!)

Here is the basic timeline of the game:

  • Feb 15, 2021 – Steam “coming soon” page launch
  • EA Release Date: Feb 28, 2021 (again, minimum duration allowed) – Only Korean Was Supported.
  • August, 2021 – English translation added. ( At this point it only had 32 reviews)
  • August 3rd, 2021 – First major coverage by English-language Youtuber. You should watch this because you can hear how he describes the game: “Terminus is NEO Scavenger meets Project Zomboid with a dose of Cataclysm” – Note Look at how he uses X meets Y! Everyone loves X meets Y.
  • December 2021 – Very famous Korean Youtuber plays the game
  • June 2024 – Daily Deal – “We ran a Daily Deal in June, which provided great exposure on Steam ahead of the full release”
  • 1.0 Release Date: Aug 19, 2024

Early Access launch

I asked In-geon what was in the game at the EA launch:

“When the game launched in Early Access, it had just one of the current six endings—reaching the Terminus—and none of the special zombies or many of the crafting and cooking recipes that are now available. It also only supported Korean at that time. If you’d like to see what it looked like back then, check out Yippee Ki Yay Mr Falcon’s video, which shows the game about five months after the EA release.

In the first couple of years, I didn’t focus much on sales. There were too many aspects of the game that needed improvement, so I devoted myself entirely to fixing bugs and adding content. I only started paying serious attention to marketing around 2023.”

In-geon

Analysis

The main reason Terminus: Zombie Survivors had such a soft EA launch was because In-geon just didn’t market it. It didn’t have many wishlists, the Steam page was up for only 2 weeks, and he didn’t participate in Steam Next Fest or any other festival. Also the game only supported the Korean language.

Despite the rush to EA, the game still had an ending so players could play through at least one character path.

Streamer-friendly game

YouTube was a far larger source of traffic than Twitter. Besides the video I mentioned earlier, Splattercatgaming covered our game four times during Early Access, which was a huge help for reaching markets outside Korea.

In-geon

After the EA launch and the English translation, the game entered a 3 year long stretch of constant updates and streamer coverage. 

Analysis

There is a common refrain among marketing that you don’t want to over-expose your game. Don’t show the demo too early to Streamers because once they play it, they will never play it again. 

But if the game is “Streamer-Friendly” they will play it over and over again. Splattercat loved the game. He basically checked in on it every single year to see how it evolved over time:

Splattercat:

This streamer played the game for 8 hours and 19 minutes straight and still didn’t run out of content! He even played it several more times (as recently as 2 weeks ago).

I don’t agree with the advice that you should hold back from promoting your game to Streamers. If your game is highly replayable, Streamers will not get tired of it. This fear of overexposure is the perfect example of falling for Loss Aversion. If you have something that is decent and not too buggy or janky, send it out. It is better to have the chance that early streamers play the game and you get that visibility, than to wait and miss out on it forever. Even if a streamer plays it and says it isn’t good, now you know! You can fix it. You can reach out to streamers later and tell them you fixed the problems. 

The upside of a Streamer playing your game way before launch is so much higher than the downside of them playing it early and not liking it. 

Discount Strategy

I place a high priority on ensuring that players don’t regret their purchase, which is why I didn’t discount the game at all for the first few months. However, since not offering discounts wasn’t drawing in enough players, I decided to go with the minimum discount of 20% that triggers Steam’s email notifications. I didn’t want to offer a deeper discount during Early Access and risk disappointing those who supported the game’s growth from the start.

In-geon

Here is the discount chart for Terminus. As you can see, no discounts for the first 10 months of development.

Advice:

Steam was built around the idea that you will discount your game. Discount at every opportunity you can. I know Factorio doesn’t discount, but you are not Factorio, and there is no way to test the counterfactual if they had used a traditional discounting strategy.

Fanatical bundle

In August of 2022 Terminus: Zombie Survivors was featured in a Fanatical bundle.

The Fanatical bundle was helpful in the end, but looking back, it wasn’t the ideal approach. Some users viewed the inclusion in a bundle negatively, and I regret that.

In-geon

Analysis

Several years ago bundles such as Humble and Fanatical were a huge way to get visibility on your game. But today I think it is mainly for bargain shoppers on don’t impact the bottom line. As you can see from the game’s Steamdb follower chart, the bundle (A) didn’t have a major impact on visibility. The curve continues upward despite the bundle. 

My gut tells me that these bundle platforms are best used for your back catalog and then using that visibility to cross promote to your newest game. It doesn’t really work if you are mid development for a single game. 

The 1.0 Launch

I started thinking about the 1.0 release last year. Once I had added all 15 planned occupations in version 0.9.8 and introduced the first non-Terminus escape ending—Escape by boat—later that year, I felt it was time to prepare for the full release.

In-geon

One month before the 1.0 launch, Terminus had collected around 60,000 wishlists. The momentum was really building. In-geon did a real marketing campaign for this version of the game: influencer marketing, paid ads, Keymailer masthead, press releases, and more.

I usually recommend against paying for streamers because their coverage typically falls flat but In-geon used a really good technique of paying only for tested streamers

I reached out to influencers who had previously played Terminus or Project Zomboid, offering sponsorships for videos. This approach was quite helpful in generating initial sales.

In-geon

Here is one of those paid streams

The slow build of the following paid off. The game was quickly featured in New & Trending. Also many many streamers played it upon launch. 

Marketing a Korean Game

In-geon is Korean and the game got its early foothold by marketing exclusively to that audience. I asked him what the major social platforms are for the Korean market:

While it might be fun to try and promote your game on these sites, note that Terminus didn’t really take off until the English translation and the English streamers started to play it. So don’t expect a huge impact from marketing it to Korean game players here.

So how did Terminus pull it off?

As my data scrape showed, 99% of games that don’t succeed within the first month fail to sell well upon 1.0 launch. So why did this game eventually take off?

In short, I think the lack of marketing and the fact it was Korean-only, and that the game was never discounted meant that the low-sales at EA launch were not indicative of the quality of the game. Basically at EA launch the game basically didn’t exist. It was untested.

You can see that as soon as the game was translated to English a major Youtuber picked it up and loved it. That is a good sign that there is something there. 

Similarly, as the game was updated streamers like Splattercat kept coming back to play it. That is a good sign.

The core game was good. It just needed content, and marketing. 

But Chris you said this wouldn’t work

Yes, Terminus: Zombie Survivors succeeded despite not following the current “meta” or “best practices” of Steam.

However, that doesn’t mean…

LOL, nothing matters, marketing doesn’t matter, do whatever you want

That one guy on reddit

No, please don’t take that lesson from this. You should still follow the best practice of 6+ months of gathering wishlists, Steam Next Fest, 7000+ wishlists, Popular Upcoming. 

I am in awe of In-geon’s fantastic game and I commend him for what he has done. However, with all due respect, I think his success could be even greater if he had followed best practices. 

Imagine if back in February 2021 on the early days of his “coming soon” page he did playtests to gather feedback and build wishlists instead of going right into his EA launch. 

Then, what if he did Early Access launch after all these were met: the English translation, and at least 7000 wishlists collected, and Steam Next Fest. By doing this, I have no doubt that the hockey-stick growth we see today at the 1.0 launch would have arrived sooner during a theoretical Early Access launch in 2021.

With a hit much earlier he could have done at least 2 Daily Deals / Weekend Deals per year which would have fueled even more visibility and more sales. Then his 1.0 launch would start at an even higher level and bring in even more sales than we see today.

Again, I outline this alternative universe not because I want to cast doubt on what is already amazing success. I do this because I don’t want you (my dear reader) to misinterpret this success as a sign that you can ignore best practices. You must account for the huge opportunity cost by not following the meta. 

Summary

When I was in University, the scariest thing in my study of computer science was “The Halting Problem.” Here is the full definition but in short it says that there are computational problems that are so complex that we can never know whether the computer will eventually solve it after grinding on it for a very very long time, or that the problem is just an impossible task that will send the computer looping forever and ever. I get this lonely and mystical feeling in my gut when trying to think about infinity. I don’t like it. 

I am so happy for In-geon and his game Terminus: Zombie Survivors but its success inspires the same existential dread I feel when thinking about The Halting Problem. For any game that does poorly in its Early Access launch, is it wise to keep grinding on the game, improving it, building a community and eventually 1.0 with it. Or is it better just to wrap the project up, do a quicker 1.0 launch, and then start work on a second game. 

We can never know. 

For 99% of games, they never succeed from a weak EA launch. But there is always a chance you could have a Terminus: Zombie Survivors on your hands. 

I asked In-geon for his advice:

Since Early Access, I’ve kept in touch with a few players who’ve logged hundreds of hours in the game. They’ve been a tremendous help, not only by providing feedback, reporting bugs, and suggesting ideas, but also by offering encouragement to keep the project going. If there was any early sign of success, it came from people like them.

Honestly, if my current level of success had been the goal back then, I might have felt discouraged. But at the time, I was a solo developer, and by 2023, I was making enough to get by (I don’t have high living expenses). By then, I was thinking, “This isn’t so bad—I could keep doing this forever.”

We can never know which game will be a hit with the 1.0 launch but here are my thoughts of signs that you might have a game that will eventually do well:

  • You are making an endlessly replayable “crafty-buildy” game. These games are what EA was built for. If you have a short, linear game it probably won’t improve by 1.0. 
  • When you launched you hadn’t done any marketing and are therefore are “unproven” game. If, on the other hand, you did a huge marketing roll out, contacting streamers, submitting your trailer, gathered thousands of wishlists, and it still underperformed, that game has been tested by the market and probably indicates a fundamental issue with the game itself.
  • You have a dedicated set of followers that play for hundreds of hours and are constantly giving feedback.
  • You have streamers who play the game over and over (See Splattercat who would play the game again every couple of months)
  • Long median playtimes greater than 2 or 3 hours.
  • Your playerbase and review count and concurrent player count continues to grow month after month.
  • You have a low burn rate (low cost of living and few employees) and can sustain development forever.

Look at this vginsights concurrent graph for the game. You can really see the line creeping up continuously. A good sign that things were in store for the game.