Marketing PC Games is so 2005, Gross.

How you can hot-wire mobile to sell PC games profitably at scale

Note: If you need assistance publishing your game, I’ve founded a new independent game publisher Ex Populus.

In mobile, it’s not who you know, it’s what you know. Don’t have any connections at Apple or Google? Doesn’t matter. The most successful publishers in mobile leverage a type of performance marketing called paid user acquisition.

User acquisition is a method of advertising where the advertiser’s goal is to buy users at a price that is less than the revenue they are likely to generate from the user. The primary variables used to determine efficacy of an app or campaign is LTV:CAC, where LTV means Life Time Value and CAC represents Customer Acquisition Cost. If a mobile marketer has an LTV that exceeds their CAC, they can scale while profiting.

This is the business model used by the biggest publishers in mobile gaming. User Acquisition is different from traditional brand advertising where advertisers goals aren’t per unit ROI, but instead to spread general awareness.

Interestingly, user acquisition is not a common practice at scale for PC or console games. If you speak with marketers in the space, they don’t really offer any specifics as to why UA doesn’t work for PC, but instead give a Dothraki-like folk explanation “It is known”.

You see, PC gaming is still stuck on an older “brand” style advertising model:

• PC and console game publishers leverage advertising, but campaigns usually take the shape of brand ads that lack measurement on a per unit basis.
• One reason is that Steam and the consoles don’t provide API access to the conversion data needed by marketers to build a data supported funnel with attribution. Marketers are blind.
• As a result most sales on PC and Console are driven by platform features, discounts and influencers. None of those things are nearly as predictable or repeatable as UA.

This creates operational issues for game developers looking to scale:

• PC is an environment where the ability to scale is less in the hands of the developer or publisher and more in the hands of the platforms (PC and Consoles) and streaming ecosystem (influencers).
• Platform features and influencers as marketing levers aren’t predictable, consistent or repeatable and often rely on the good graces of a human.
• This makes establishing predictable unit economics more inconsistent, nondeterministic and unscientific.

On PC and Console, it’s not what you know, it’s who you know.

The Dothraki were fun to watch on TV, but let’s not base our growth model on human impulse or accept the folk legend as it stands. In order to solve this problem, we need to deconstruct the mobile ecosystem along with the primary sales channel for PC (Steam). In this process, we will identify key hurdles to run a successful paid user acquisition campaign for a PC game. Once we understand each key hurdle, we can build a solution that directly addresses each hurdle.

Device Proliferation and Screen Time

The first problem selling a digital product targeted to PC or Console users with a performance model is device proliferation and screen time. We live in a time now where almost everyone has a smartphone and spends much more time looking at the small screen.

As a result of this, publishers can widen the top of their acquisition funnel by marketing to users on mobile devices instead of waiting for them to sit down at their gaming PC. Since we are marketing to mobile devices, we will also need to ensure the sale happens on a mobile device.

Problems to solve:

1. Avoid marketing to PCs and Consoles.
2. Optimize towards marketing to people on their smartphones.
3. Don’t break the chain by expecting the user to switch devices.

User Experience and Funnel

If you’ve purchased a game on Steam and downloaded a game on the AppStore, consider the difference in the user experience / funnel. Everytime a potential customer is asked to do something to continue, it’s additional friction and a new conversion point that needs to be overcome. Each conversion point will have significant drop off as a result of the potential customer deciding not to continue.

On the AppStore:

• Tap the ad
• Tap the get button
• FaceID or passcode

On Steam:

• Click the ad
• Enter birthdate
• Scroll to find purchase button
• Click to purchase
• Self or gift?
• login (more than a click)
• Enter verification code (more than a click)
• Press proceed to Steam button if code accepted
• Enter cc # or at least the security code (more than a click, get wallet!)
• Agree to terms
• Press purchase

In the above charts, we simply use a 50% conversion rate at every event to illustrate the point that Steam has a terrible (impossible) funnel (actual benchmarks will probably be worse). It’s no wonder paid user acquisition doesn’t work as well on Steam. Steam’s user acquisition funnel has 11 steps (one which you may need to physically get up from your chair) compared to the AppStore’s 3 steps. Expect CAC on Steam to be exponentially higher than it is on the Appstore simply because there are so many hurdles. Advertisers need to spend dramatically more to overcome the conversion decay, which makes running performance ads untenable.

We also need to consider the page layouts. It is hard to illustrate in the layout funnel above, but look at these screenshots of Steam’s game pages. The colors and visual design are fine, but the layout is atrocious in terms of building a page to sell.

In the above image, I simply navigated to a Steam game page on my laptop. Where is the purchase button? Well below the fold. We also see over a dozen links above the fold leading me away from this page to pages where I won’t buy things, or to my competitor’s games. This is terrible, but the mobile site is even worse:

On the mobile site, not only are we missing an above-the-fold purchase button, but the primary media used to sell the game is also partly below the fold. This is atrocious. Compare it to Apple’s game page, where all best practices are followed:

Apple shows you an icon, image, ratings, title, tagline and most importantly the get / buy button all perfectly fit to your screen without the need to scroll.

Summary of problems to solve:

1. Significantly reduce conversion points in the funnel and therefore friction.
2. Improve sales page layout dramatically
3. Don’t make consumers go get their credit card.

Attribution and Tracking

The final issue is a bit more “behind-the-scenes”. Paid user acquisition requires events to be setup throughout the funnel so data can be sent programmatically in real-time to user acquisition platforms like Facebook and Google. We need to tell Facebook and Google (anonymously) when the user they sent clicks or buys. The ad platforms use this event data to optimize and find “look-a-like” audiences. Facebook and Google know that their systems need to target users that “look like” the ones that are clicking deeper into the funnel and ultimately buying your product. In mobile, companies can set up events in their own apps and games leveraging measurement partners like AppsFlyer.

Unfortunately, Steam’s 11 conversion points there is no way to setup events to send the ad platforms since no one can get access to Steam’s page code and there is currently no API.

Publishers would also need to have something pinging the ad platforms from within their game with a Unique Identifier to notify the ad platforms that a new player / sale was a result of the campaign. In mobile the ad id (IDFA) is used for this purpose. There isn’t really a windows equivalent.

Summary of problems to solve:

1. Conversion Event Tracking.
2. UID for unique sale/customer/user identification.

SOLUTIONS

Let’s do a quick review of all of the problems we need to solve to have a proper paid user acquisition operation for PC games:

1. Avoid marketing to PCs and consoles.
2. Optimize towards marketing to people on their smartphones.
3. Don’t break the chain by expecting the user to switch devices.
4. Significantly reduce conversion points in the funnel and therefore resistance.
5. Improve sales page layout dramatically
6. Don’t make consumers go get their credit card.
7. Event tracking.
8. UID for unique sale identification

The advantage that we have in selling PC games is that putting a game on Steam and the consoles is a best practice, but you can still sell your game directly from your own website. This isn’t something mobile marketers can do because Apple and Google have closed systems that do not allow selling apps out of the stores on websites.

Solutions (1,2,3): Avoid PCs, Market to smartphones, Don’t break the chain

We don’t want to deal with hosting our own game file and dealing with touchy subjects like DRM. To address this will be selling keys. Steam and the consoles make this easy, which is great because we are still driving our users into Steam’s ecosystem which comes with plenty of social components and other benefits.

https://partner.steamgames.com/doc/features/keys

Because we are selling keys, we can easily focus on selling to consumers while they are on their mobile devices. Once they make their purchase, we email them the key and they can setup and install later when it’s convenient. This means we can market exclusively to gamers while they are on their mobile devices and we don’t need to worry about them needing to switch to their PC to make the purchase. The point of this exercise is the sale of premium games, so if they don’t play until later, it’s not a CAC or LTV problem.

Solutions (4,5): Reduce friction and conversion points while improving layout

Because we are selling keys to mobile users, we can easily make our own mobile optimized website. All we need is a single page that fits on a mobile screen with a buy button and some media above the fold. A trailer, gif, or even playable html5 demo will show the potential customer what they can expect. I took 15 mins to throw together a mockup:

In the above mock-up, every element is unique and original, but it may feel similar to either Steam or the Appstore. Because we own the page, we can easily and quickly split test media, colors, and other visual elements along with the advertising creative we are running on the UA platforms.

Solution (6): Don’t make customers grab their credit cards

Because we are marketing on mobile devices this problem solves itself. We can leverage Apple Pay and Google Pay. The adoption rates (80%+ in the USA) are so high, it’s unlikely the users will ever need to go grab their wallet. Just tap to pay with the credit card already setup on the device. It’s relatively simple to detect if the user is on iOS or Android and show them Apple or Google Pay respectively.

Solution (7): Event tracking and attribution

Event tracking is solved because we own the page and can set up any cookies or click events we want. We have all the data so leveraging it to optimize look-a-like targeting on Facebook and Google is a given. This section is the shortest because it’s the simplest answer, but this simple answer is probably solving 50% of the problem. Attribution and data is core to UA.

Solution (8): Unique sale or customer UID

This issue has been solved by Gamesight.io which is a thriving startup that has made an attribution platform for PC and console games utilizing digital fingerprinting as a substitute for the IDFA. With Gamesight we can properly communicate to the ad platforms when they have driven a unique sale with Gamesight’s unique sale id so they can further optimize look alikes around unique sales data.

Result on direct sold:

• Tap the ad
• Tap the checkout button
• Tap pay
• Face id or passcode

Now that we have our solutions, let’s build a quick proto-type and watch the sexiness in action:

Appstore: 3 Steps

Direct sold: 4 Steps

Steam: 10 steps

Publishers may possibly eliminate the “checkout” step for direct sold given that users are required to verify the purchase with their passcode or faceid with Apple/Google pay. The solution above I amalgamated in an hour with off-the-shelf tech. If so, the process would be as lightweight as the Appstore, with event tracking and targeting optimization every step of the way.

Because we have eliminated most of the friction, setup proper data and event tracking while targeting the most convenient devices, we should be able to get UA economics similar to mobile.

With the above solution, publishers can still put their games on Steam and the consoles, but they’ve created a new channel and lever that allows them to leverage Facebook and Google, the largest platforms on earth to create ROI.

Additional Benefits

With this solution we own all the data and user experience. This means publishers can leverage:

Data management platforms

Proper DMPs like Blue Kai to optimize targeting and acquisition economics. Because we can purchase demographic cohorts from BlueKai and map them to our Facebook and Google campaigns, our optimization cycles will be reduced, and so will our CAC.

Affiliate Networks

Now, because we have such a highly performant sales page and own all the data, we can connect to the APIs of the biggest affiliate networks in the advertising space. We can introduce bounties to other marketers that can bring us new users. These bounties are in line with our LTV economics so profitability is insured. We are still using Gamesight so we know each sale is new and unique.

Shopify Ecosystem.

We could always integrate with the payment providers directly, but leveraging Shopify is a no-brainer since it is a mature ecosystem that comes with a number of benefits. For example, it is directly tied into influencers on Instagram, Twitch and plenty of others. There is an entire app ecosystem around Shopify, so possibilities are high.

Discounts, Giveaways and Sweeps

Now that we have our own one page mini-store, we can discount, run give-aways and even sweeps. These can be great lead capturing tools for those not ready to purchase. They can’t enter to win for free and we can market to them later.

It’s still Steam

Because keys are being sold, all sales are being funneled into Steam’s ecosystem. None of the benefits of Steam are lost, but all the weaknesses are removed.

LIKELY QUESTIONS:

Why aren’t other big studios doing this?

Many PC games have terrible direct sale experiences. Their sites lack obvious optimization to reduce friction and convert sales. No mobile channel focus. Ridiculously long funnels, poor page layouts with important bits below the fold and likely no event or sale attribution. This is because PC has evolved around brand style advertising and until recently there haven’t been attribution tools available to make this style of marketing efficacious. Most publishers focus on the big platforms, where attribution like this is still impossible.

Many mobile marketers that understand how to sell with performance campaigns are making their way to PC/Console via mergers and acquisitions, but they are coming from an environment where direct selling is made impossible by Apple / Google, so it is a new way of thinking for them too.

This approach may only be effective for self-publishing developers. If you are a pure publisher, you automatically have a smaller LTV because you are only taking 30% of the revenue. To optimize for LTV>CAC business model, developers should self publish their own games to find the best economics, which is common in mobile.

Distributors (like Steam) take 30% off the top. The remaining 70% is split 70/30 again in the developer’s favor, which is illustrated above. As you can see the publisher that is in charge of marketing must find a CAC that is less than their cut. For a $20 game that means the CAC must be less than $4.20. For a healthy margin, a publisher must sell their game for 2–3 dollars a sale. Very difficult for a $20 price point.

However, if the developer is self publishing, they are selling against a $14 LTV (on a $20 game) after the distributor takes their cut in the worst case. In Steam’s key documentation, it appears Steam does not take a cut of direct key sales. In this case, developers have even more room because the entire $20 is theirs (perhaps less a standard processing fee). I won’t bother with another pie chart showing what 100% looks like, LOL. In theory, you can spend $20 to drive a sale and make your money back on organic and lift from curated sales.

If you are reading this and thinking about making a new store, platform or publishing business to make all of this work, keep in mind the economics. This is best leveraged by marketing savvy game developers. If you’re just trying to “get a piece” your piece may not be big enough.

Big companies move slow and indies in this space are generally the starving artist archetype versus the innovative marketing guru found in startups in other slices of tech. Despite this, I would expect to see more of this type of solution soon, as it reduces dependency on human impulse and allows devs to create scale with science. If I were Epic, Steam or similar, I’d be thinking about marketing urls following these mobile and experience best practices along with API access to conversion events for developers selling on their stores. Having millions in ad budgets from other company’s coffers pointed at your store urls can’t be a bad thing.

Most mobile games are free, how does the premium price tag influence the funnel?

Given that we’ve solved the issues with friction, page layout and attribution, this is more comparable to performance marketing for eCommerce vs freemium mobile funnels. With freemium mobile games, marketers rely on the game’s loop to create a hook incentivizing players to spend money to make progress after they have already installed the free game.

With premium PC / Console, marketers must leverage social proof (ratings, likes, reviews) along with great ad creative to push the sale (in addition to the pages of best practices outlined in this article). Marketers that have worked on promoting crowdfunding campaigns, physical consumer goods and similar will have the experience to find the right conversion economics for pay-in-advance models for games. I’ve heard good things about Metric Theory, New Engen and Adquadrant, but I haven’t run any test campaigns yet, so I can’t fully endorse. Those are probably good places to investigate if you need a performance team.

Don’t free mobile games have a larger LTV?

Most of the time they don’t. This 2018 report (Pg. 15) that was created from a collaboration between Appsflyer and Facebook indicates that the average LTV for free mobile games is about $1.70 on day 90. Including only paying users, the LTV increases to $70+, but this is irrelevant because we can’t pretend the acquisition cost of non-paying users isn’t coming out of our bank account and increasing the average CAC!

While social casinos have the highest LTV at a reported $300+, those LTVs are also calculated by excluding the users that never pay (most). When freemium mobile publishers run UA campaigns, more than 95% of their users never pay. Those costs can’t be excluded in this exercise. On the other end of the genre spectrum is hyper casual, which is primarily ad monetized. These games typically see LTVs below $1.00.

If you’re selling a premium game on Steam, 100% of players have paid. So if the game is $20, the LTV is $20 (as a developer selling keys directly). This $20 LTV is comparable to the initial metric of $1.70 LTV. In most cases premium PC LTV > freemium mobile LTV.

What about the key sites, isn’t this what they do?

Selling keys was one part of this proposed solution. The other elements are not addressed by key sites. From what I see there is no mobile focus, no funnel optimization and no attribution. Key sites are just additional store fronts like Steam with too many purchase steps, and too little access to conversion event data to do paid UA effectively. Further, they dilute the LTV by taking a cut.

OK, so eventually everyone else will start doing this too, won’t they?

Great! The barrier to entry to this type of marketing is tiny. The hard parts are the targeting, creative and split test optimizations that need to happen once you kick off. Of course you’ll need a game to sell and a budget to sell it with. :)

This document illustrates best practice. Game creators should sell their own games on their own html, leverage the data and minimize conversion points to create ROI positive unit economics with UA style performance campaigns. Existing third party tools like Shopify, BluCommerce, Sendowl, Squarespace and Wordpress make this simple. This is a mobile, performance mindset coming to PC game sales. Developers take note.

Forget the laptop, just buy it on your phone

Author’s Ethos

Tobias Batton is a 15 year digital product & marketing executive that has founded 3 companies and sold 2. Most recently, Batton was responsible for over 150 million app installs delivered to top mobile game publishers with his Signal Zero publishing and loyalty platform. In 2012, He was hired by the Emerge Digital Group to build a mobile division for the organization which helped the Emerge be named America’s 8th fastest growing company in the USA (Inc. 500). Prior to that Batton was tapped to build the real time event tracking system and temporal search engine Live Matrix, which was adopted by Facebook to build Facebook Live.

Batton was also a founding team member of IGN’s Indie Open House which helped indy game developers create and release their games for desktop and console. Batton also founded Resistor where a variety of popular web and mobile games were published including the widely played Clan Wars and the cultural phenomenon iGirl. Batton’s first company which helped independent filmmakers publish their films digitally was launched in 2006 and acquired in 2008.

He was responsible for millions of R&D dollars building a loss prevention platform for digital advertising while managing several millions in quarterly marketing budgets. As a result, over 100MM installs were driven to top mobile game publishers while simultaneously preventing hundreds of millions of dollars worth of loss for ad technology companies that represent billions in annual market share. He has led the development of a mobile SSP, built a mobile promotional platform with 60MM+ users and was a part of the launch of thousands of games and apps. Because of his work promoting games, he was featured on Fox Business Risk and Reward. He has guest lectured doctoral candidates from Stanford, MIT and INSEAD on the above topics. He is intimately familiar with mobile devices, information theory, Shannon entropy, device fingerprinting and a variety of other related concepts.

Batton has designed and published several best selling mobile and social games. This led to his published game design framework “Variable Dopamine- Wavelength Mapping Theory” — a system based on Dopamine spikes in the brain, applied directly to an overlap of both Maslow’s pyramid and Pavlov’s trigger-based framework — has been featured at numerous conferences and articles, including Forbes. He has been profiled by Forbes, Fox Business Network, the New York Times, WIRED, IGN, Cnet and VentureBeat.