As Mark Twain never said, history doesn't repeat itself, but it does rhyme.

I used to get so many pitches from startups eager to knock PC gaming powerhouse Steam off its block, that in 2018 I wrote one big standard response called So You Want to Compete with Steam, with a follow-up a year later. The dust has now settled and the result is clear: all of the new contenders* failed but Epic, and even they have a long upward climb ahead of them.


*For clarity I never considered Itch.io and GOG.com as "new" contenders -- they've long since carved out a comfortable and steady niche for themselves, but do not represent any kind of major threat to Steam itself

Flash forward to today, and my inbox is stuffed with pitches from start-ups wanting to compete with Roblox, that plucky Lego-ish multiplayer game-creation platform currently valued at 41 billion dollars.

So I guess we're gonna do this again.

Here's how you can build a successful business that competes directly with Roblox:

Don't!


I say this out of love: the vast majority of you are going to fail. I admire you and your hard work and dedication; I'm pessimistic simply because your task is incredibly hard.

First of all, you are late to this party. Roblox first launched in 2006, a full fifteen years ago – that's five years before Minecraft, if you can believe it. They have a massive head start and are playing by an entirely different set of rules. Your only chance is to flip the entire problem on its head.

Honestly, re-reading my old articles on Steam I'm feeling guilty for not being harsher with everyone. I know of at least a dozen companies who went bust or gave up, and many of the same challenges from last time apply here. My sworn duty is therefore to put the fear of God into your heart so that you wake up and do one of three sensible things:

  1. Quit now
  2. Pivot to a beautiful blue ocean instead
  3. Take this extremely seriously and leave nothing on the table

If you're still reading this, congratulations, you are bold and/or crazy enough to have selected option 3. The deck is massively stacked against you, but best of luck anyways.

Minimum Requirements to not Fail Right Away

You should consider the following as mere table stakes:

  • High quality multiplayer support for user creations out of the box
  • High performance servers with excellent reliability
  • Powerful, user friendly, and joyful creation tools

What Roblox and Minecraft have in common is all their user generated content is natively multiplayer. Roblox's key edge, however, is that it's really easy to make and share interesting content, whereas with Minecraft mod creation is an unnecessarily technical pain in the butt. Do not underestimate the importance of high quality and easily accessible tools.

These are just the minimum requirements, by the way. They're essential to get right, and way harder to pull off than most people give them credit for. Doing them all perfectly simply grants you a spot on the starting line.

With that accomplished, let's move on.

The Typical Pitch

A dead giveaway of a doomed startup is when their pitch copies Roblox, and then deploys exactly one of the following strategies like it's some kind of secret weapon:

  • "Our platform is like Roblox, but is technically superior!"
  • "Our platform is like Roblox, but with better graphics!"
  • "Our platform is like Roblox, but takes less in revenue share!"
  • "Our platform is like Roblox, but will do <quirky thing>!"
  • "Our platform is like Roblox, but for when kids age out and want something mature!"
  • "Our platform is like Roblox, but for the educational market!"
  • "Our platform is like Roblox, but runs in a browser!"

None of these are bad ideas –  in fact many are very good ideas – but they're not sufficient by themselves, not to mention that everyone else has already thought of them and are deploying them right now.

Addressing each one briefly:

Nobody cares if your platform is "technically superior" unless that manifests as something customers and developers actually care about, like better performance, better usability, or awesome features. Worse, "beating Roblox on features" is way harder than you think, because an App is a collection of tiny details and Roblox has had 15 years to noodle on those. You likely can't match them feature-for-feature and probably shouldn't even try because at best you'll just wind up exactly like Roblox but without an audience and nobody will care. You should focus your technical prowess on things Roblox doesn't or can't do.

Better graphics seem like a win, but could actually be a liability. An intentionally crappy aesthetic is actually a strength Roblox shares with Minecraft – not only does it make it faster and cheaper to render, it keeps young content creators from feeling self-conscious that their contributions aren't "good enough" or "real" when contrasted with everything else on the platform. Also, Roblox's standard blocky avatar allows for standardized cosmetics, one of their chief monetization schemes.

Better revenue share is good, but this goes without saying – of course you can't charge as much as the incumbent! More royalties are irrelevant without a sizable audience; content creators know that 100% of zero is still zero.

Nobody cares about your quirky gimmick unless it manifests in a unique experience they can't get anywhere else, and don't underestimate the ability of established platforms to just copy it should you start to show traction with it.

Counting on kids to age out of Roblox and move on to yours is a big assumption that might not be true – have you researched whether mass-market audiences age out of specific games, or do they simply age out of playing games entirely? Or do the games themselves travel alongside the generation so "aging out" never quite happens and they're still playing Pokémon 20 years later? I don't have all the answers, but on PC at least I've seen evidence of huge numbers of players sticking with a single game for years and years and never playing anything else, ever. Are you sure enough about this "aging out" phenomenon to stake your entire business on it? And even if kids genuinely do grow out of Roblox, who's to say Roblox won't provide a more "mature" offering themselves?

If you're going for the educational market, beware. My career started in educational games and I can tell you from experience it's a brutal, toxic place. I could write an entire article on that alone so here's the highlights – 1) if you're working in academia proper you will be forever chasing committees and approvals and grants while juggling petty collegial politics and back-stabbing, 2) if you're working in the school system your game will be shoved through a standards compliance colonoscopy procedure with incredibly detailed and arbitrary learning goals and curricula, 3) the entire space is drenched in the principal-agent problem, 4) most people in education know nothing about game design but will insert themselves anyways, 5) most people in the games industry know nothing about pedagogy but will insert themselves anyways, and 6) your product had better run on whatever ancient, low-spec, locked-down computing hardware the institutions happen to have on hand. Companies in the educational games industry are often mistaken for lightweights by the "regular" games industry, but they're actually extremophile species specifically adapted to navigating this strange world.

What about running in a browser? As I detailed in The Future of Games is an Instant Flash to the Past, this one could actually be a key feature for you. It lets you do a complete end run around the app stores, and lowers friction for users and content creators alike. However, it's technically demanding to get full browser compatibility across mobile and desktop, and not every "browser-based" solution is created equal. Remote cloud streaming, for instance, has limitations that client-based solutions don't – players must log in before they can play, and you're stuck with a 1:1 ratio between peak player counts and the number of gaming machines in your data center.

Speaking of cloud, let's touch briefly on a common pitfall:

The Cloud Cost Trap

An acquaintance recently reminded me that many startups misapply Jeff Bezos' famous quote "your margin is my opportunity." Yes, you can and should charge less than the incumbents' super-high fees, but you first need to grasp the cost structure of scaling this sort of business.

Look at Roblox's purported revenue share breakdown:

Source: gamerefinery.com

Roblox is being disingenuous when they claim to take "24.5%" here, clearly their share of this particular graph is 50.5%. But the point remains that Roblox spends a lot of money on hosting and network infrastructure (here's their S-1 for more details) even though they're a mature platform, and if you're not careful you will wind up spending far more than they do per user.

This classic problem happens when companies start off using commodity cloud technologies like AWS and Digital Ocean, and once they're operating at scale a few years later they suddenly realize cloud provider costs eat up a huge portion of their revenue. At that point "repatriating" their service onto their own servers is an expensive and painful multi-year transition. The trouble is that when you're still a startup you often can't afford the fixed time and overhead costs of setting up your own servers. Threading this needle can only be overcome with careful planning and experience and you need to get it right before it becomes a problem.

If you screw this up you'll have no margin left for yourself, let alone enough to undercut your competitors' revenue share. And don't forget that the specter of burning money for years on user acquisition and creator incentives still looms.

Okay, let's say you have a good idea or two and all the necessary technical wisdom and basic business sense to get the job done and avoid the immediate up-front pitfalls. That doesn't make you special, it just keeps you from false starting and disqualifying yourself before the race even begins.

Now let's move on to the real problems.

The Real Problems

The big three problems you're facing are:

  • Chicken-or-the-egg deadlocks
  • Platform dynamics
  • Ownership and trust

Let's start with the chicken or the egg.

Chicken or the Egg Deadlocks

One of the key themes of So You Want to Compete With Steam was a nasty paradox best articulated in Joel Spolsky's Strategy Letter II: Chicken and Egg problems, which also applies to would-be Roblox competitors:

  • You need players
  • Players won't show up without content, so you need creators
  • Creators won't show up until you have players

Joel points out that you can't expect this deadlock to solve itself – instead you need to just go out there and deliver a truckload of chickens or a truckload of eggs. Typically this means spending a lot of money. Anyone able to rely on organic growth alone started ages ago and that door is now closed to you.

I cannot overstate the importance of seriously grappling with this problem. Don't just stick your head in the sand and noodle on fun technical features.

Content creators value audience size above nearly everything else, closely tied with fat stacks of cash money. It doesn't matter how easy it is for creators on your platform to make mind-blowing experiences if you don't also have an audience eagerly waiting to play them.

Seed your platform with awesome material by paying your own employees to build beautiful creations. Hire contractors and independent content creators and then pay your staff to train them in your tools. Pay these people to make tutorials and guides and videos and post them all over the internet and don't stop. Set up an affiliate system with creator and influencer rewards. And that's just the obvious stuff – you need to be thinking about new and innovative solutions to this problem 24/7. Pay any and every price to get high quality content onto your platform.

Once that's done, use all that great content to attract players. Just don't make the mistake of being "generous" at the expense of your creators – if you give out any premium creator works as freebies to players, the creator had better get paid out of your own pocket just as if a customer had bought it.

Even after you've done all this, you're still going to face all the usual user acquisition cost ramps to get players to sign up. Good thing you came prepared with truckloads of money!

That said, there is one weird trick that can even the odds.

The Power of Instant

If your platform has the natural advantage of running in a browser – an "instant game" as the investor-types are calling them – you have some incredible advantages.

First and foremost, Roblox itself does not run in a browser. This is a rare genuine leg up you can get on the king. I would be shocked if Roblox hasn't at least tried to get a web build up and running, but it's a safe guess they're working with a codebase whose roots extend as far back as 2006, with layer upon layer of upgrades and rewrites piled on since. You're nimble, they're not.

Second, running in a browser makes the top of your user acquisition funnel wider, because all it costs to try your platform is a click. No downloading an executable, no installing anything, and no App store. Just please, for the love of all that's holy, don't throw your advantage away by making guest users register and sign into an account before they can play. If a user can just click and be playing seconds later you will enjoy entirely new possibilities for your game to spread via social media, chat apps, long-winded blog posts like this one, and things you never imagined.

Imagine if instead of this dumb meme, the coolest game on your platform was just running right here on this page, right now:

Why not make your games as easy to share as this stupid meme?

How many cool platforms have you read about but never actually tried? Well, if your platform was truly instant, everyone who reads this blog post could have just experienced it for themselves. (No, don't email me asking if I'll embed your game in this blog post. Do it yourself).

Finally, running in a browser lets you cut the app stores and their rent-seeking behavior out of your life and your bottom line. Claw back that 30% for yourself and your creators, while every other platform – including Roblox – is still hobbled by App store taxes.

Next, let's talk about Platform dynamics.

Platform Dynamics

What do we even mean by the word "platform?"

For the purpose of this article I use "platform" to mean a digital ecosystem whose owner gets other people ("creators") to build stuff for players to enjoy, and the owner makes money off of this activity somehow.

At least within the context of Roblox-adjacent games & interactive multimedia, a  typical "platform" consists of 5 things:

  • Creation tools
  • Playback engine
  • Discovery methods
  • Marketplace / transaction engine
  • Relationship with the customer

Not all platforms are the same. As much as I hate to quote Bill Gates positively on anything, he was absolutely right when he said this about Facebook:

This isn't a platform. A platform is when the economic value of everybody that uses it, exceeds the value of the company that creates it. Then it's a platform.

This is a great insight, but I'm gonna modify it a bit – instead of pedantically quibbling about whether platform X is a "true" platform or not, I prefer to grade platforms on a spectrum from open to closed:

Platforms graded on a spectrum of "Open" to "Closed"

On one end you have open platforms like the World Wide Web where each of the five aspects is owned by no one but the commons.

Towards the middle you have different kinds of closed platforms like Windows and Steam where certain components of the stack are proprietary, but others are unowned; the owner either refrains from (or is simply unable) to capture most of the value that creators produce on the platform.

On the far end are digital company towns, proprietary platform stacks privately owned from top to bottom. In the physical world company towns are communities where a single corporation is not only the sole or principal employer, but also owns all the housing and stores – the company is your boss, your landlord, and even your grocer. Total ownership grants the company power over not only every aspect of their workers' lives, but also their families and the entire local economy. Digital company towns likewise squeeze as much value out of creators as possible.

Let's look at a few well-known platforms and see how they break down under this model:

Web Windows Steam Flash Roblox
Creation Tools Unowned Unowned Unowned Owned Owned
Playback Engine Unowned Owned Unowned Owned Owned
Discovery Methods Unowned Unowned Owned* Unowned Owned*
Marketplace Unowned Unowned Owned Unowned Owned
Customer contact info Unowned Unowned Owned Unowned Owned
Customer financial relationship Unowned Unowned Owned Unowned Owned
Kind of Platform Open Closed Closed Closed Company Town

The Web is an open platform. You can create content with a variety of tools both open source and proprietary, publish to an open format based on open standards (CSS/HTML/JS), render and execute the content in a variety of competing web browsers, and connect to and transact with customers from all over the world using any variety of independent financial services, with no single authority able to mediate all of your activities and charge rent.

Windows is a closed platform mostly in the sense that software written for Windows generally won't run on other platforms (cross-platform tech aside). Despite being only one aspect of the platform stack, this one is a huge deal because resource constraints force many developers to commit to one platform exclusively. Developers are stuck with the native system API's, but can use whatever creation tools they want, and customers can find, purchase, and install Windows-compatible software from any source without necessarily involving Microsoft. This isn't because of Windows' benevolence, mind you – if Gates had had his way he would have strangled the Web in its infancy, and with each new release we see signs of Microsoft trying to extend its tentacles into formerly unowned domains.

Steam is also a closed platform. Like Windows, you're not restricted to using special Valve-owned tools just to make games, but unlike Windows you don't need to use a Valve-owned executable format just for Steam to be able to run your games. Sure there are some proprietary Steam API's, but they're optional. The principal thing Valve actually owns is the marketplace and the relationship with the customer. They also own Steam's on-platform discovery, but I've noted this with an asterisk* because no platform is ever able to totally enclose the act of discovery, nor should they want to – or else there would be no games journalists, Youtubers, Twitch streamers, TikTok influencers, etc.

And let's not forget Flash, dead as it is.

Flash was a closed platform in the sense that Adobe owned the tools (Flash IDE) and the playback engine (Flash player/browser plugin). But they failed to capture the rest of the ecosystem, leaving the community to build it for themselves. This stemmed not from Adobe's benevolence but from their incompetence. As a tools-focused company they just didn't get platform dynamics, and made bizarre efforts like trying to charge for "premium" API's that never worked. The Flash ecosystem famously thrived in large part because of Adobe's failure to completely enclose the platform. (I covered this in The Future of Games is an Instant Flash to the Past if you want to read more).

As for Roblox...

Roblox is a Company Town

Roblox games must be created with Roblox Studio, can only be played in the Roblox App, are chiefly discovered in the Roblox discovery portal, and can only be monetized with Robux through the Roblox Marketplace, with all customer contact and financial information retained by Roblox itself. This gives Roblox all kinds of natural advantages, from resisting competition to locking in their audience, as well as preventing their content creators from gaining any negotiating leverage.

I have seen several pitches already from platforms who naively expect to be able to replicate the same company town model. They'll own the tools and the playback engine and the marketplace and the relationship with the customer. Some of them even think they'll be able to charge 30%!

Look, I know some of you as customers actually like company towns from giant companies like Apple precisely because they're locked down and you trust the platform holder. Good for you, sincerely! You are more than welcome to continue liking them as a customer. But this article isn't addressed to you; it's addressed to startups who think they can deploy this kind of vertically integrated stack without already starting from a position of strength.

Simply put, if you're trying to build a Roblox competitor in 2021 under the company town model, you're delusional. You should not build a company town for two very good reasons:

  1. Company towns are bad, and you shouldn't do bad things
  2. It's way way way too late to succeed with this strategy

Roblox and a few also-rans can get away with a completely closed stack top to bottom because they started long ago. They already have a built-in audience. You need to convince people to try your platform while all the established platforms and the recent flood of competitors nip at your heels.

So what can you do?

You Must Give Something Away


Give people a reason to build on your platform. Make them owners, not tenants.

What should you give away? Well, that depends on your specific situation, but I recommend "as much as you possibly can." Recall the five components of a platform:

  • Creation tools
  • Playback engine
  • Discovery methods
  • Marketplace / transaction engine
  • Relationship with the customer

As much as I hate Microsoft and Bill Gates, if you think about it it's amazing how little of the the total economic value generated on the Windows platform is actually captured by Microsoft. Had Windows been conceived of in 2015 rather than 1985 I would expect it to refuse to run any program that Microsoft hadn't specifically approved of and to force all sales to go through an official app store taking 30% of everything and jealously seizing the customer relationship for itself. It's shocking how modern platforms have come to see this as some kind of divine right.

Platform taxes impose deadweight lossif Apple takes 30% of my revenue, and AWS takes another 25% (or whatever), then there's a lot of things I might have considered doing with my business that just aren't profitable to even consider. Apple and Amazon get rich, but for every dollar they make, many other dollars of potential economic value are annihilated.

Just as Land Reform spurs innovation by turning tenant farmers into entrepreneurial owners who are finally allowed to realize the gains from their own hard work, investment, and innovation, giving content creators ownership over the tools, playback engine, discovery, marketplace, and even the customer relationship itself incentivizes them to settle down and make cool stuff on your platform.

This brings me to trust.

Ownership and Trust

Platforms tend to follow a certain kind of life cycle, and there's no better primer than Dan Cook's Game of Platform Power. In it he outlines how platforms transition through "Growth" and "Engage" phases where they are friendly and generous to the creators who produce value on their ecosystems, before maturing into the "Extract" phase where they leverage their size and power to lock-in users and capture as much creator-produced value for themselves as possible.

Once upon a time, platforms said things like "Don't be Evil" and "users will never be required to use a Facebook account to use an Oculus" and we've seen what that's worth. Take YouTubers and Twitch streamers for example – they're constantly demonetized and copyright-struck in bad faith and just generally taken for granted by the platforms. This despite the fact that content creators have made Youtube as valuable as Netflix without Youtube having to spend a dime on original programming.

No matter how generous your platform is today, content creators aren't dumb, they know how this works, and they're being exploited right now by company towns like Roblox. Words are cheap. What they want is assurance. Trustless assurance. And no, I'm not talking about blockchain.

You really want to shake things up?

Give content creators a loaded gun pointed at your platform's head.

Another word for this is "exit rights." If you want creators to come over in the first place, give them the power to leave anytime they want.

Exit Rights

What do Substack, Supercast, and Ghost have in common? All of these platforms give content creators full ownership over the customer relationship. And I'm not just talking email addresses. I'm talking about the full financial relationship – these platforms give you full control over a Stripe account where your customers transact. They're subscribing to you, not to the platform. These platforms take their share, but only as long as they're providing you with value. Should you decide they're no longer holding up their end of the bargain you can walk away and take all of your customers with you in the blink of an eye.

By conventional wisdom, this is madness. Why would you make it easy for your customers to leave? Well, believe it or not customers and creators are human beings, and are motivated by carrots at least as much as by sticks and locked dungeons. Provide a good service and they'll probably stick around.

And don't underestimate the threat these unassuming little companies pose to established media platforms. In actuality they are relentless digital crowbars laser-calibrated to pry talent away from existing platforms. And should you scoff, just look at how quickly and how hard these companies were able to hit Vox and the New York Times. Or look at the example of The Hill, a news site now frantically seeking a buyer in the wake of the departure of its massively popular morning show hosts who are now running their operation independently, powered by Supercast.

To be clear, I'm not here to cheerlead for any particular startup or platform, just to trace the shifting outlines of platform power.

Case in point: Substack was hit with controversy right away as the linked NYT article above describes in detail, leading some of its newly signed writers to pull the trigger and leave shortly after joining up in the first place. But whether you like Substack or hate it, those departures ironically strengthened its position in my view, because they were dramatic and highly visible proof that Substack's touted exit rights are real. The writers who left Substack departed for other technically similar platforms like Ghost that they felt were less objectionable, but still offered the same kind of exit rights.

As an aside, this blog is hosted on Ghost, but that's an unrelated coincidence – I set this thing up ages ago before Ghost pivoted to be a Substack competitor.

Exit rights aren't just about "being nice." If you circle back to platform dynamics, you'll realize that giving content creators true ownership of something encourages them to settle down, plant roots, and start making stuff. If they're just working for a landlord the edge of their passion will be blunted, but give them a plot that's truly theirs and they'll do amazing things. The more you squeeze, the less oxygen that's left, and the less productive the entire platform will be.

One of the most valuable things you can give away is the customer relationship. As for other things, yeah you can outsource discovery, but since you can never fully enclose that in the first place it's not much of a giveaway and will probably just be seen as a sign of laziness. That leaves creation tools and your playback engine. Are you bold enough to make your tools and runtime open source and publish your games in an open format? Most startups I've talked too are reluctant to give up control over their creation tools and playback engine in particular, seeing it as their "secret sauce." Maybe so, but if you don't give away something you're back to competing with a dozen other companies on pumping out features and seeing who can shovel money into the user acquisition furnace the fastest.

So here's a few other reasons to consider giving away tools and runtime platform.

Tools are Hard

It is not easy to make good tools. And not just for the normal software engineering reasons – ie, you'll inevitably become too familiar with them and become unable to see their flaws, you'll forget to dogfood them and build them around speculative users rather than actual things creators want, etc. No, tools are hard because they take time. Not just time to build, refine, debug, and test, but time for creators to learn, adopt, and build them into their lives and workflow. This is much harder than a typical user acquisition problem for a simple game or CRUD app, it's more like inventing a new instrument and expecting musicians to start writing and recording music for it.

Two things in particular make creators wary of adopting new tools: 1) being afraid they'll suddenly disappear, and 2) not having ownership over what they create with them. If the only thing your tools are good for is making games and experiences that run exclusively on your platform, and could disappear entirely if your company goes bust, that limits their appeal. You're a startup and history suggests it is extremely likely you will go poof. Exit rights in tools are a trustless assurance to invest in your platform despite that risk.

The trouble is that opening your tools isn't all that valuable without also opening your runtime platform. If people can export all their asset models and scene graphs, but can't actually bring them to life without your platform, their value is limited to novelties like 3D printing their avatars and buildings. And if you open source your playback engine, you do run the risk of someone cloning your service and competing against you with your own code. On the other hand, you have a chance at making your technology become a standard, with your platform at the center of it.

So You Want to Build the Metaverse

Another buzzword you see thrown around constantly is: "We're building the metaverse."

For those of you who don't know what the "Metaverse" is, it's a 3D virtual persistent space that's meant as a successor to the internet itself. The term originally comes from the book Snow Crash, but has since been popularized and reimagined in movies like The Matrix and Ready Player One.

The key thing about "Metaverse" startups is that they're not just meant as games platforms, but all sorts of experiences – a cool weird magical place to just hang out and do anything. The aspect a lot of startups miss when they insist that their Robloxified platform is "the Metaverse" is that the Metaverse isn't something that any one person can or should own, it's the entire connected space. "Closed platform" and "Metaverse" is a contradiction in terms.

This sort of massively ambitious project needs widespread buy-in, spontaneous participation, and unplanned innovation, all on an enormous scale. That's at least if you're sticking to the original vision of an actual successor to the internet and world wide web as a massively persistent embodied space. I think that's pretty far fetched, but if you scale it down all you're doing is redefining "Metaverse" to mean "My little multiplayer UGC app," kinda like how we redefined "Cyberspace" down to just a synonym for "The Internet."

That's not to say someone fundamentally can't craft a "Dark Metaverse" under the company town model. It's just that their name is Facebook, it will be a dystopian hellhole, and you don't have a chance of competing on those terms.

If you really want your platform to become the seed for "The Metaverse" then you need to give it away.

Good Luck, You'll Need It

Look, I know this all sounds a bit crazy, but you're the one who decided to compete with Roblox. I'm just here to make you fully aware of the magnitude of the problem, coming up with an exact business model that threads that needle and adapts it to your specific situation is your job. The entire venture is crazy to start with, and my point is simply that the riskiest thing to do is play it safe.

If you do something wild and ambitious and off the wall you might still fail but at least you'll have stood a chance. Even better, you'll greatly increase the odds you'll discover a weird new opportunity along the way and pivot away from "competing" with Roblox, accomplishing something much better instead.