Disastrous Game Launches Are Avoidable

push to talk #36 // on concord, deadlock, and the importance of market signal

I’m hesitant to write about Concord. I’ve worked on launches that went wrong, so I know the pain those devs are feeling. My heart hurts for that team.

But the unprecedented nature of Sony’s decision to pull Concord from store shelves this week raises the stakes. The scale of the debacle—of the time and money lost—demands a deeper look. Anyone serious about marketing games is obligated to face the situation and try to figure out what we can learn from it.

That’s the theme of this week’s newsletter.

---

Last Friday, IGN reported—quoting friend-of-the-newsletter Simon Carless of GameDiscoverCo—that Concord only sold about 25,000 copies in its first week.

This, everyone agreed, was a catastrophic failure. Only days later, PlayStation announced its plan to kill the game and refund players. The internet quickly piled on to proffer theories for Why Concord Failed.

But most of these explanations left me feeling deeply dissatisfied:

• In the IGN piece, a bunch of level-headed, smart, and buttoned up business people blamed a combination of low awareness, weak retail promotion, and market saturation for shooters (which all makes sense, although shooters with no marketing budget or retail support do 100x better on Steam every day)

• It’s because it wasn’t free-to-play, others insisted (despite the fact that successful F2P launches—especially of new IPs—are more rare than ever)

• It was the cinematic trailer that “doomed” it, claimed one YouTuber (lol… no?)

• A lot of people blamed the character designs (I can’t find the source now but someone said the game looks like the live-action adaptation of a different, pre-existing animated series)

• And of course, at least one “game dev” on Twitter who hasn’t shipped a game in 20 years blamed it on woke (because for culture war grifters, every issue is a culture war issue, and also please follow me on twitter! please!! follow me!!! care about me!!!! i was team lead on a very popular game that released in 2004 but have failed to achieve anything of note since then so now you need to sign this petition and like my posts so i can feel relevant to this industry!!!!!)¹

And, I dunno man. I just don’t think any of these really explain how a very high budget game published by one of the world’s most powerful brands failed to hit 700 concurrent players and got outsold by Squirrel With A Gun.²

Jason Schreier’s quip that Concord had “no hook or strong word of mouth” comes closest to a satisfying summary, but even that’s more of a description of observable fact than an analysis of root causes. Tom Warren at The Verge came in hotter, saying the game was “worse than bad—it was forgettable.”

So I’ve been asking friends around the industry for their takes—the kind of thing they wouldn’t say on the record.

A friend who is very senior at a major platform holder had this to say:

Hero shooters need to be good at some things, genre-defining even. This is a red ocean game so player expectations are higher.

Characters are one of those things. They made theirs detractive. Apex also has detractive heroes. But the gameplay is awesome.

But Concord has uninspiring gameplay that doesn’t do anything remarkable or new. So together, coupled with no marketing… that is bad.

The character designs tell me two things:

1) They didn’t talk to any players in a meaningful way.

2) Their design culture had either no ability to dissent or no ability to hear dissent.

That is more of an indictment than the characters themselves.

I don’t know if this friend is right. As far as I know, they’re not operating with inside information.

But they raise a great point: Wasn’t there some way to find out that Concord had a problem before the game released?

---

The Warning Signs

I first started paying attention to Concord about three months ago and immediately got kinda worried for the dev team.

The game had the opening slot in PlayStation’s State of Play presentation on May 30th, and comments from players on YouTube were brutal. Not just negative in a general sense, but extremely cutting and focused on a couple of consistent points.

The first was the idea that the game looked like Overwatch but worse:

Then there was the perception that the game’s world and characters weren’t creatively distinct. Guardians of the Galaxy got mentioned a lot:

You could also definitely see some culture war-style backlash against specific character designs. This one confused me at first:

It turns out there’s a trope in right-wing communities that the phrase “updated for a modern audience” is shorthand for game and/or Hollywood studios reworking existing IPs to—I guess—make all the boobs smaller (and less polygonal?) and replace white characters with a more diverse cast.

Apparently this is a hot topic for 12-year-olds and emotionally stunted adults, but as a grown man with the ability to get along with people who don’t look like me, I really don’t give a damn that they made The Little Mermaid black.³ It’s usually wise to ignore the algorithmically boosted weeping of the internet’s most maladapted posters.

But the culture war crap only made up a slice of the response to Concord. More concerning was the reaction to Concord’s positioning against other games in the market, and the fact that almost nobody seemed to be clicking with the characters. Where was the fan art? The stans? To put this in boring marketing terms, the game wasn’t connecting with any significant audience segment.

That’s why in early July—6 weeks before Concord’s launch and 2 weeks before its open beta on Steam—I linked to that Concord trailer in a newsletter with the throwaway comment that “if you find yourself getting cooked in the YT comments section you can probably expect the same negative reaction on other platforms.”

I thought the game was in trouble. But I didn’t think it’d go this badly.

The Concord open beta was the first real shocker. It was basically a free-to-play weekend for the game on Steam, and it peaked at only 2,388 concurrent players.⁴ I usually caution against using CCU charts as the ultimate measure of a game’s worth but… that is an extremely low figure for an open beta of a big-budget game on Steam.

And the numbers started dropping immediately. It had no word-of-mouth growth, and those that tried it were not sticking to the game. By Sunday, the open beta’s final day, the game barely escaped triple digits with its peak CCU: 1,199. Paul Tassi at Forbes called it correctly when he wrote that “There Is No Spinning How Poorly Sony’s ‘Concord’ Open Beta Went On Steam.” If the game flopped this hard as a free download, how could anyone expect it to do well as a $40 title?

But by this point, it was too late for the Concord team to change course. The PS5 discs had probably already been printed. And the game had reportedly been in development for eight years already. How much could delaying it by a few months really change things?

So here’s a thought experiment: How might things have been different if the game had been shown to players earlier? Like—much earlier?

What if Firewalk Studios had done an NDAed, 5,000-player private playtest of Concord a year ago? Or two or three years ago? What might they have learned?

Almost certainly, they would have seen how quickly players bounced off the game. They would have gotten brutally honest feedback, with enough time to rethink the launch altogether.

For multiplayer games that need to survive as a live service, there’s really no excuse for waiting until launch to find out that your game is a flop. You can learn that and a million other things by showing a game to players sooner. Especially if you’re attached to a brand like PlayStation, you have the resources to get people in and validate your core assumptions.

There’d be the risk of leaks, of course, but who cares? Players have gotten used to seeing games in early development. They’re capable of seeing past the limitations of an early build and imagining the future potential. So why are developers of multiplayer games waiting until launch day to show their games to lots of players?

The alternate path, which the creators of Concord could have taken, is to be much more open earlier in the process, and unashamed of giving the public access to WIP content. If you’re planting seeds you expect to grow for years on end, why not stick your head outside and check the weather first?

There are ways to get real signal from the market earlier. And you need that signal, because the market is ultimately going to decide your fate.

The prime example for this alt approach is Valve’s handling of Deadlock.

---

It Looks Like Butt and Plays Like a Dream

Just a multiplayer game in early development. No big deal! Thanks Valve.

Let’s talk about Valve’s go-to-market strategy on Deadlock.

I think most people have been just bemused by the way Gaben and Co. have been halfway pretending the game doesn’t exist, despite the fact it’s now one of the ten most-played games on Steam. If you go to its (only recently made public) Steam page, there’s hardly anything—just a little warning box that says…

Which, like… okay. That doesn’t actually tell me About This Game!

To learn about Deadlock, you have to get an invite and play it, like millions of other people who are falling in love with it. This part is going quite well for Valve. The game is growing every week and popular streamers are already announcing that they’re switching their focus to Deadlock.

The nontroversy around The Verge’s “violation” of the game’s easily-dismissable non-binding NDA was also extremely funny. But it begs the question: why exactly is Valve doing it this way?

Here’s my theory: The Valve team learned from their experience with Artifact and Underlords. They learned that not even the world’s greatest PC gaming company is immune from releasing a dud every once in a while.

So with Deadlock, they’re seeking market signal early. They think they’ve got a great game on their hands, and they’re almost certainly right. But they didn’t want to wait around to find out. Based on the length of time they’ve run this test and the way they’re limiting access, it’s clear they’re trying to get a sense of how the game retains and grows on its own, with as little marketing support as possible. They’re trying to get market signal.

I’m saying market signal here instead of player feedback for a reason. Player feedback can tell you what specific players think and say about a game. Market signal tells you how players (writ large) might actually interact with it. And right now the signal on Deadlock is insanely strong. People I work with are saying they could play this game for 10,000 hours while still enjoying it and learning more about it. We’ll see how that plays out in reality. But if the game really does turn out to be that sticky, it’s great news that Valve will have learned way, way before the game’s “official” launch, whenever that may be.

But what if it had gone differently? Deadlock could have been dead-on-arrival like Artifact, or it could turn out to be a great game that nonetheless steadily sheds players until it dwindles to near-nothing, like Dota Underlords. There are many such cases. The only way to get a sense of a game’s fate is by showing it to lots of players. And not waiting until the final hour to do it.

What’s a better way to get market signal: doing tiny private playtests with a greenlight committee, or simply putting a game out there to see how players react to it? The only thing stopping you is your outdated assumption about how “brand value” works. (The truth is, no one cares.)

At least for multiplayer games, there are plenty of ways to get market signal before launch, if you’re willing to seek it. And, hell, you might actually have some fun in the process.

---

That’s it for this week. I’m gonna go spend 18 hours learning the new wallride techs in Deadlock.

I’ll see you next Friday.

1. Sorry, got a little carried away there.

2. I mean, to be fair, being a squirrel with a gun sounds hella sick.

3. However I will be boycotting Snow White. Not because they made Snow White latina, but because those dwarves look absolutely haunted.

4. The SteamDB chart for the Concord beta app is here.