The Wages of Cheap

Unity’s CEO, John Riccitiello has some ideas about what it takes to succeed in games and chose to express them in a really careless way. General sentiment towards Unity was falling at least among some developers due to shifting priorities, and a recent round of layoffs. The latest outrage came not so much from an interview (which went largely unreported at the time) but from subsequent reporting on that interview with more inflammatory headlines.

Riccitiello’s comments are not particularly interesting, save for the fact that he is the CEO of a widely used game engine, and a lot of people have been very outspoken about it. There are some interesting areas of discussion, but these sadly seem to have been overtaken by an undue emphasis on the less substantial parts of the interview. This article will look at the lesser appreciated aspects of Unity’s latest announcement and whether or not developers actually need to do anything in response.

What did Riccitiello say?

Because so much of the attention has been placed on the quote that made it to the headlines, it is probably best to get it out of the way before addressing the substance. Here is the full quote as printed in the pocketgamer.biz interview:

Ferrari and some of the other high-end car manufacturers still use clay and carving knives. It’s a very small portion of the gaming industry that works that way, and some of these people are my favourite people in the world to fight with – they’re the most beautiful and pure, brilliant people. They’re also some of the biggest fucking idiots.

I’ve been in the gaming industry longer than most anybody – getting to the grey hair and all that. It used to be the case that developers would throw their game over the wall to the publicist and sales force with literally no interaction beforehand. That model is baked into the philosophy of a lot of artforms and medium, and it’s one I am deeply respectful of; I know their dedication and care.

But this industry divides people between those who still hold to that philosophy and those who massively embrace how to figure out what makes a successful product. And I don’t know a successful artist anywhere that doesn’t care about what their player thinks. This is where this cycle of feedback comes back, and they can choose to ignore it. But to choose to not know it at all is not a great call.

I’ve seen great games fail because they tuned their compulsion loop to two minutes when it should have been an hour. Sometimes, you wouldn’t even notice the product difference between a massive success and tremendous fail, but for this tuning and what it does to the attrition rate. There isn’t a developer on the planet that wouldn’t want that knowledge.

The initial article’s headline was: “John Riccitiello: this industry divides people who shun monetization and those who embrace what makes a successful product.” This is in contrast to mobilegamer.biz’s reporting on the article the next day: “Devs not baking monetization into the creative process are ‘fucking idiots’, says Unity’s John Riccitiello.” Neither headline misrepresents Riccitiello’s quote, but it’s easy to see why the second headline caught broader attention.

What was he thinking(!)?

While the provocative framing of later headlines has not served the discourse well, it did at least draw attention to the comments. I think Riccitiello knew what he was doing and was trying to be provocative. The whole passage reads like he’s trying just a little too hard to be cool. He’s already sung the praises of creatives, so he gets to trash them and use profanity because he’s the big CEO who calls it like it is and has “… been in the gaming industry longer than most anybody…”. It is something of a bold strategy for the CEO of EA from 2007-2013 to be bragging about their time in the gaming industry, but nobody is accusing him of lacking confidence.

This is probably where the discussion should have ended. CEO throws muck, developers throw back, everyone feels like they should take a bath and promises to be better in the future. This is best thought of as the “what was he thinking!?” question (a question of mental state), where a CEO embarrasses himself at a particularity bad time. However, a lot of responses focus on the wrongness of his emphasis on a more service driven model of development. This is the “what what was he thinking?” question (a question of content) concerning the state of the games industry from an influential CEO.

Some may find this a distinction without a difference since Riccitiello’s views on the future of gaming are as offensive as his efforts to look cool, but there is a cogency to his vision of gaming. He is saying that monetization needs to be considered earlier in a game’s development, and that Unity’s latest (and controversial) acquisition is about enabling that. There is room for disagreement with these views, but it is not clear that the content of his comments are disqualifying so much as the tone.

Is he right?

Riccitiello’s vision of gaming is not unassailable. Dividing between developers that “…throw their game over the wall…” and “…those who massively embrace how to figure out what makes a successful product” is a false dichotomy. But this can be dealt with by simply asking: do developers need to embrace the more service oriented model he is proposing in order to succeed?

To see the sense in this point of view, consider the state of the world. It is widely accepted that the majority of indie games released on Steam do not make a profit (Ryan Clark’s video on hooks puts the rate of profitable indies as low as 5%). Jeff Vogel goes so far as to argue that there are too many games and it represents a massive waste of human effort. In contrast, Riccitiello’s view is optimistic. To him, the problem is that these games did not appropriately consider their market, and if they had had better information they would have succeeded.

Relative optimism is not going to be persuasive if the concern is an overly commercial view of gaming. Indeed, some may even be averse to thinking of games as a commercial product at all (falling into the “throws it over the wall” stereotype). The commercial side should at least be acknowledged, as developers have a responsibility to someone to at least break even on a game. Investors and lenders expect some kind of return, and it is inappropriate to take money without the intention of repaying it. Employees are going to want some kind of stability, which requires at least enough financial success to meet payroll on time. Some developers may be self-financed or are able to operate due to the support of loved ones, but these developers should not be setting terms for what the industry should do.

Recognizing commercial imperatives does not imply adopting a full blown service model for every game, but consider two alternatives for such a game. One is a game that is developed under the traditional model that Riccitiello criticizes, while the other has considered its service model from inception. The first game is much more likely to feel like the monetization elements are bolted on at the last minute and is more likely to annoy or even exploit the player. The second game has had the benefit of time to think things through and integrate monetization more seamlessly into the game. Bad people will continue to do bad things, but a rushed and uninformed approach is far more likely to have good people do bad things, if only in error.

Even in an idealized case, however, the service model is highly contestable. Developers who are more artistically focused also care about what the player wants, but do not necessarily think that the kind of information provided by Unity’s analytics tools provide the insights they need. EA’s own attempt at adopting these tools during Riccitiello’s tenure did not pay off.

Henry Ford likely did not say “if I asked the customer what they wanted, they would have said a faster horse that ate less” but the sentiment has certainly been expressed by successful manufacturers and creative professionals. Ken Kocienda’s book Creative Selection: inside Apple’s design process during the golden age of Steve Jobs describes a creative process free of focus groups and analytics for products that are undeniably successful and have had an influence few others can match. In contrast, Amazon’s highly data driven approach has struggled to produce the kind of success in gaming that makes the company feared in nearly every other sector it touches. Not only is there evidence that these kinds of analytics are not necessary for the creation of successful products, there is at least some evidence that this approach has gotten in the way of developing successful products and has produced unnecessary disruptions to the overall business in the form of layoffs and disappointing sales results.

However, the fact remains that many developers and publishers do maintain teams of analysts to understand how the games are performing and what their players are doing. The games using these tools are financially and critically successful, making it hopeless to try and draw the kind of clean split reactions to the interview would like to make between art and commerce. Even among independent developers, the tendency is to prefer to be informed. Indie devs did not spontaneously decide to prioritize wishlists and rediscover demos. These trends came about from data driven articles on what it takes to succeed on Steam.

Why does Unity think this way about games?

If a particular vision of gaming is going to be proven right, it will likely be done in the marketplace, and so a more interesting question may be why Unity has chosen this particular vision. Unity began as a game developer but ultimately decided to switch to selling its tools instead of making games. The company still describes itself in terms of gaming, and its largest source of revenue comes from gaming. As tempting as it is to reduce unpopular decisions down to whims or indifference, Unity’s actions are that of an experienced team executing on a specific strategy based on their understanding of the industry. That strategy may be in error, or they may fail in execution, but Unity’s motives are rarely those we first think of when frustrated with the company.

As a public company, Unity’s views on its performance and strategy are available in its regulatory filings. The annual report is called the 10-K and it contains financial results as well as a discussion of those results. Under the ‘key metrics’ section (page 66), one measure is the number of customers who spend more than $100,00 on their services (this number has grown from 600 to 1052 over 2019-2021). Further down, Unity breaks down its sources of revenue as create solutions: subscriptions for Unity’s products, operate solutions: things like Unity’s ads and other services, and a category for strategic partnerships and other revenues. As a share of revenue, create solutions have been more or less stable (31% to 29%), while operate solutions have grown from 54% to 64% over the same time period (2019-2021. The difference is made up from a fall in strategic partnerships). In fact, one of the risks Unity identifies earlier in the report is that over half of its revenue comes from operate solutions.

If it wasn’t clear already, Unity does not primarily make its money from its engine. The engine still accounts for a third of what Unity sells, which translates to hundreds of millions of dollars, but it is neither the largest, nor the fastest growing segment. Unity’s discussion of the risk from the share of operate solutions does not discuss mitigating that risk through diversification, but rather the need to attract and retain customers. In fact, the next key metric is ‘dollar-based net expansion rate’ which tracks how much more or less existing customers are spending.

An analogy might be to think about an old (pre-Alphabet) iteration of Google. To users it’s a search engine, to investors it’s an ad company. Unity’s users are more inclined to think of it as the application on their desktops, but to understand the actions of Unity as company it is Unity Ads, deltaDNA, Multiplay, and all those other icons cluttering up the product’s page when you are looking to download the personal edition.

This perspective on Unity is not new. Unity has often boasted about how over half of all games are made using its engine, and how billions of players interact with these games. These numbers cannot exist without mobile (Unity currently claims 72% of the top 1,000 mobile games are built with Unity). Mobile relies on the kind of services that form Unity’s largest source of revenue. So far this approach has been very successful, and developers have benefited. The present model does not require Unity to raise subscription fees in order to achieve profitability, and so users are able to use a high-quality tool for little upfront cost.

Whether one looks at Unity’s key metrics, or how it prices its services, it is fair to say that Unity considers its primary customer to be someone who spends more than $100,000, and that, while the engine is a necessary first step to a successful game, that success will require additional services that the company is happy to charge for. This approach has succeeded for them in the past and so very little of this is news. Ultimately Unity, and those it considers its primary customers, have succeeded more under a service based model, and its decisions reflect that. The service games can be thought of subsidizing the tools for the more traditional games that don’t use Unity’s operate solutions.

Is the merger any good?

The catalyst for all the discussion was Unity announcing a merger with ironSource. In addition to the controversy regarding John Riccitiello’s comments, concerns were raised regarding one of ironSource’s past products and the decision to do further acquisitions after recent layoffs. These factors have led people to conclude that the merger is a bad decision for Unity, and the 14% drop in share price as an indicator that financial markets agree that the merger is a bad idea.

Whether or not the merger is a good idea, the drop in share price has very little to do with the merger. In so far as price movements are informative, it is as an indicator of investor sentiment, since buyers and sellers have a financial incentive to analyze the company closely and trade on their best understanding of the company’s prospects. This view serves as a first approximation, and it should be remembered that the consensus valued Unity at $196.65 a share in November and $33.71 a share 8 months later in July. The factors listed above (developer alienation due to the comments, ironSource’s past, layoffs) may be things that affect the financial health of the company, but the share price went down on the 13th, while the outrage surrounding the interview arrived a day later.

Unity made multiple announcements on the 13th. One was the merger with ironSource, another was that it was lowering its earnings guidance for the overall year (that is, Unity communicated that they think they will make less money than previously thought). Earnings are much more closely tied to estimates of the company’s value, and so such an announcement will predictably cause the share price to fall. While the market may have thoughts on the merger, they are impossible to disentangle from the response to lowered guidance.

However, there are reasons for concern, even if they are not tied to share prices. This first major issue to address is ironSource’s label as a “malware company.” This label is not the most accurate characterization, though the facts are still damning. The reputation concerns a product called installCore which was classified as a ‘potentially unwanted program’ which, at least under some definitions, is distinct from malware in the sense that the former is annoying, while the latter is intentionally harmful. The strongest case for malware (beyond choosing a definition to suit a conclusion) came from a blog that pointed out the program could be switched to install malicious software, but there does not appear to be any instance of doing so. This clarification is not intended to excuse the program. installCore, by all accounts, was very bad! Installers should not put unwanted programs on your computer, and they should not try to trick you. Not only was the program bad enough to be blocked by Windows Defender and Malwarebytes, the program apparently checked to see if certain antivirus programs were present before reaching out to install the extra software.

Unity attributes the problems to “bad actors” and highlights the fact that the program has not been offered for a long time (the product was cancelled in 2020). Customers have long memories, and so it is possible that ironSource has reformed and is still being penalized for its past. Unfortunately there’s no way for us to know if the circumstances or people that led ironSource to produce a product like installCore are still in place. Unity needs to maintain its relationships with platforms (a selling feature for the engine), and, even if there are problems prior to the merger that are waiting to be discovered, it will be Unity’s reputation that is tarnished. Even if Unity’s monetization efforts simply become more obnoxious, the outreach numbers it likes to boast about can be expected to fall. The difficulty is that attractive acquisition targets often have some undesirable feature that makes the acquisition viable and so it can be tricky to distinguish between a savvy move and a catastrophic waste of money.

There is a deeper concern regarding the merger, and it regards the layoffs. Unity’s recent layoffs were concentrated in the artificial intelligence and software engineering segments, suggesting that the company may have found it easier to scrap its own monetization efforts and simply buy someone else’s. If so, it raises serious questions about the overall management of the company, especially in light of internal complaints about “Strategic pivots at a rapid, unpredictable rate.”

The concerns with the merger deepen based on another announcement Unity made on the 13th regarding its financing. Unity is acquiring ironSource by giving shares in the combined company to ironSource shareholders. Unity’s own share price is near its all time low, meaning more shares of the company are being given up than would have been in the past. It is difficult to see what Unity might have done otherwise as it does not directly control the share price, does not have enough cash on hand to do the acquisition without stock, and ironSource shareholders might have accepted a lower price given their participation in any upside from the combined company. One concern that emerges from such deals is the dilution of existing shareholders, although it should be noted that, in principle, dilution should not be a problem if the acquisition is fairly valued, since shareholders are giving up part of the company in exchange for something of value (so, more people sharing a bigger pie. If Unity overpaid then more people share a not quite as big pie, meaning some people have less than before). It is what Unity announced to deal with dilution that is concerning.

Along with the other announcements, Unity announced share buybacks. Just as companies can issue new shares to raise money, they can buy back shares, offsetting the dilution that occurs from an event like the ironSource merger, or generally raising the share price (same pie shared by fewer people). The buyback program is for up to $2.5 billion, and lasts for 2 years, effective after the merger closes.

Finance can get complicated quickly, and the tricks of running a public company may not be of interest to someone who is just annoyed that the CEO called them an idiot, but it can be helpful to play out the logic of the merger. Unity is acquiring ironSource using stock at some of its lowest valuations, and expects this merger to be good for business (“$1 billion in adjusted [earnings] by the end of 2024”). Unity then intends to purchase shares of the company just as its prospects are rising when, presumably, the prices is higher than it is today. Unity is also taking on debt to finance these activities, borrowing $1 billion at 2% interest, with the option to convert the debt into shares (in addition to the $1.5 billion in debt it already issued, also convertible, but at 0%).

This money should be reinvested in the company instead. Unity’s rhetoric is that it is a growing company. Indeed, the only way its consistent lack of profitability makes sense is if these losses are being incurred in order to build a company that will be even bigger and more profitable than if it had not. Unity is not unusual in this regard. However, Unity went to capital markets for finance a little under two years ago, and acquired six companies, some with tenuous links to its core focus. As outlined above, the latest acquisition may be repairing operational problems in the operate solutions business. Now they are giving some of that money back. Either Unity has run out of ideas, or it is appealing to shareholders at the expense of the improvements it could be making and, given the layoffs and decreased guidance, likely needs.

What does it mean for developers?

Corporate activities do not need to be a preoccupation for developers. Discontent with Unity’s products predate its IPO, although becoming a public company certainly provided the well-worn rhetorical club of “they only care about shareholders” to anyone upset about the engine. Unity has always been accountable to those who have financed it, and neither Sequoia Capital nor nameless fund managers are likely responsible for DOTS being late and frustrating to use.

While the merger and John Riccitiello’s comments got an unusual amount of attention, there isn’t a lot of news so far as Unity’s overall approach and interests. If anything, smaller developers are simply waking up to a fact that has been true for years: they are not Unity’s primary customer. Not being a primary customer does not mean a developer is unable to use the tools, but it can result in the kind of culture shock that we are currently experiencing.

Unity did not always have such favourable licensing terms. Historically Unity used to only have the free and pro editions, with pro costing $1500 US, and addons for iOS and android costing between $400-$1500. The current engine is better than its predecessor and costs considerably less, especially after accounting for inflation. These more favourable terms have come about because of the service components that, to judge by Twitter, are anathema.

Unity’s CEO is going to want everyone to make service style games that make use of analytics and ad platforms, because that is how the company makes its money. Developers are free to ignore him, just as they are free to ignore his opinions on their intelligence.

Most importantly, none of this should be provoking moves from one engine to another. Switching is not a small matter, and even switching between projects necessarily involves the destruction of some human capital and lost productivity as people get up to speed with the new workflow. Some people already considering a switch might find this makes up their mind, but these cases reflect concerns prior to the interview. Even developers concerned about the long time viability of the company are probably better off making the most they can of the tools while they’re available. There are already far more pressing risks and complications in game development. Why would someone voluntarily make their work harder? Because the CEO dissed them? Are they going to switch to a custom engine if someone writes something nasty about pixel platformers in a pull request for Godot?

Riccitiello’s comments on development were dumb. Hopefully he gets some better PR training and doesn’t do that again. His vision for how to achieve success in gaming has served Unity up to this point, though his overall management of the company is very much an open question. Developers may not share his vision of game development, but they may still find Unity’s tools useful in following their own vision. The best course of action is really the same as meeting any boor at a party: ignore them and don’t let it spoil your evening.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s