Understanding the DOJ’s Case Against Google, What They Got Wrong, and What We Should Be Talking About Instead.
This was originally posted at tannerphilp.com, which is where most of my content goes. I would encourage you to follow me there.
On Tuesday, the U.S. Department of Justice filed an antitrust suit against Google for what it claims is anti-competitive behaviour in monopolizing search.
This post will dive into this case at a cursory level but the main thrust of my argument is that (1) the goal should be maximizing value for consumers and (2) antitrust actions like this won’t push that forward, and (3) we should be talking about how we accelerate emerging competitors, not slowing down incumbents.
This post will talk about the value of monopolies, the cost of monopolies, and what the options are — outside of antitrust actions — to deliver the most value to consumers with these in mind.
Let’s dive in.
The Google case is interesting in that the arguments made by the DOJ are simple and easy to understand, yet easily refuted.
The suit itself is long (64 pages), but if you want the jist, the DOJ was kind enough to distill their main arguments into four bullet points neatly presented in their press release. (We’ll ignore that the DOJ clearly doesn’t understand the “rule of three”, but I digress)
In particular, the Complaint alleges that Google has unlawfully maintained monopolies in search and search advertising by:
Entering into exclusivity agreements that forbid preinstallation of any competing search service.
Entering into tying and other arrangements that force preinstallation of its search applications in prime locations on mobile devices and make them undeletable, regardless of consumer preference.
Entering into long-term agreements with Apple that require Google to be the default — and de facto exclusive — general search engine on Apple’s popular Safari browser and other Apple search tools.
Generally using monopoly profits to buy preferential treatment for its search engine on devices, web browsers, and other search access points, creating a continuous and self-reinforcing cycle of monopolization.
These and other anticompetitive practices harm competition and consumers, reducing the ability of innovative new companies to develop, compete, and discipline Google’s behavior.
The key part of their argument is basically that Google has a lot of distribution and a lot of money so they pay a lot of money to put their search engine in most web and mobile browsers, and since search has the richest data-set for targeted ads ($130 billion+ / year) that compounds to more money which buys them more distribution.
The case is straight forward on its surface. The DOJ claims that Google has a monopoly position on search, which is factually correct. They claim that Google uses their monopoly position in search to make a lot of money, which is factually correct. And they claim that Google then uses that expanding revenue to pay to maintain their monopoly position, which is factually correct.
The case falls over because while the facts of the case are universally understood, Google has a simple argument for why anything other than this is bad for consumers: Google is simply better than the alternatives.
The last paragraph of Google’s opening salvo in their response perfectly articulates this:
“This lawsuit would do nothing to help consumers. To the contrary, it would artificially prop up lower-quality search alternatives, raise phone prices, and make it harder for people to get the search services they want to use”
They then proceed to compare themselves to Bing and Yahoo! search prominently (with screenshots!), which, for any reader who has ever tried Bing or Yahoo! search is effectively a knockout punch. If you haven’t used Bing or other Google competitors I’ll save you the trouble: they suck.
This is where antitrust lawsuits for software companies are a losing battle. Unlike previous subjects of large antitrust suits: Standard Oil, cable companies, phone companies, etc. that monetize consumers directly and used their monopoly positions to push out competitors and *raise* prices, software businesses like Google and Facebook monetize their consumers indirectly, so the argument for why these services occupying monopoly positions is bad for consumers is a tough sell. Their services are free, and they are getting better as they grow in dominance. As Google pointed out — forcing them to be less competitive in search would just mean that consumers have a worse experience.
There is a real argument to be made that, as consumers of these services, we should be supportive of Google, Facebook, Netflix, et al. expanding their monopoly position if we feel that an incremental dollar going to Google is going to go the farthest in improving search. Their machine might be the most efficient at innovation, so then perhaps they should get the incremental revenue to push their product forward. I mean, do we trust Google or Yahoo! to make the most of an incremental dollar?
Where this argument falls over though is that these monopolistic businesses expand beyond their core competency and use the revenue from their cash cow businesses to expand their moat in ancillary businesses. Messaging is a great example. Facebook uses their ad revenue from Facebook.com and Instagram to fund WhatsApp and Messenger which make no money. That makes it tough for other services to emerge and compete. Just take Signal as an example — that product is superior in its messaging protocol but is privately funded by Brian Acton (co-founder of WhatsApp), generates zero dollars, and therefore doesn’t have a ton of resources to innovate. That’s evidenced by the product experience (no offense Signal team — although I’m sure they would agree).
But how many other teams could build a killer experience if they had the resources and runway to execute?
The question the DOJ should be posing, and the question we should be asking, is not what is the cost to consumers, but what is the opportunity cost to consumers.
A basic principle in economics when evaluating the impact of monopolies is to look at the deadweight loss of consumption. If a monopoly is in place they set the price and the consumers who demand that product but not at the monopoly’s price do not consume the product. If that price is above the marginal cost to produce the product then there is a deadweight loss in the economy for the profitable production foregone.
This framework is generally applied to production of commodities where other producers *could* step in but they are pushed out by the monopolies.
However, this same framework can be applied to innovation. I referenced this in a blog post ~3 years ago called: Aligning Incentives in a Deflationary Economy. Here is an excerpt:
“The cost to society in a monopolistic economy is the deadweight loss, where the dominant player maximizes for producer surplus, leaving consumers on the sideline who could have participated in an efficient market. In the digital economy there is minimal marginal cost, so the deadweight loss comes in the form of lost opportunities in innovation, because a monopoly (Facebook) could take that innovation (Snapchat Stories) and implement it into their product. This further increases their monopolistic position in the market. “
Through this lens, you could look at the diagram above as the missed opportunities of potentially inventive companies. The bigger challenge here is that it’s actually less about Google, Facebook, etc. “killing” companies in the traditional sense. The likely outcome now is that these tech giants will buy a fledgling company that has some promise, deprecate the work they were doing, and put the team to work on something boring like optimizing the top-bar of the home screen for _insert_tech_monopoly’s_domain_.com.
This happens more and more than you and I would ever know. I have some friends who went through YC with different companies and I heard varying stories of the same effect that people’s goal was to “get big enough for Facebook/Google/Amazon to buy us” and then they’re set with a nice set of equity and a comfortable paycheck doing meaningless work.
See this post on Hackernews from a Google Engineer from earlier this year:
“I joined Google straight from college 6 years ago as a SWE, and by now I’m used to the style of work of “do the minimal work possible to do the job”, I never challenge myself to deeply learn about what I’m doing, it’s almost like I’ve been using only 10% of my mental capacity for work (the rest was on dating/dealing with breakups/dealing with depression/gaming/…). Even when I get a meaningful project, all I do is copy code from the internal codebase and patch things together until they work. I was promoted only once.
Now that I’m thinking of jumping ship to other interesting companies, I’m having serious doubts that I really learned what I should have learned during all those years. Especially since I’m considering companies with a higher hiring bar than Google.
How can I keep myself accountable while I’m still at the company to deeply learn the FE/BE technologies to be better prepared for other companies? Should I start by preparing a checklist of technologies and dive into each of them for a month and continue from there?”
Most people you talk to that work at the big tech companies would corroborate this statement, I promise. That’s not to say that everyone who works there is trapped in a glass case of uninteresting work and if freed they would be the next great innovator, but I am sure that a lot of the people in the “machine” could be making a bigger impact if not absorbed in the monolithic structure designed to maintain monopoly position.
So to come back to my earlier point. If I had to allocate a dollar to either Google or Yahoo! to optimize search, I’d probably go with Google. But if I had the opportunity to allocate that dollar to a promising team of hungry, smart entrepreneurs then maybe we should be finding ways to do that. Email is a good example of this. GMail has almost 2 billions users but the product is whatever. I have a lot of gripes with the product, but it works. The gripes I have are probably the same gripes you have — simple things like text entry, formatting on mobile, etc. I sometimes think that I’m taking crazy pills — does Google not see these? Do they not use their own products? No, they do see them, but it’s such a big machine that making any material changes to the app is a long, bureaucratic process and nothing gets done. I know this because I know people that *used* to work on the product but left because they couldn’t get anything done. But look at a team like Superhuman. They designed email from first principles and are building something innovative and disruptive. But Superhuman is in a bit of a unique position. The founder, Rahul Vohra, had built and sold a company previously. He has the personal credibility to get resources and the confidence to stay independent, for now. For every Superhuman there are dozens of potential disruptors that never got out of the cradle. And even with Superhuman, it’s early. I assume the Death Star is in the process of preparing to “fire when ready” (via attractive acquisition offer).
So coming back to my first point. If the argument by antitrust is that these monopolies are less good for consumers but Google can continue to swat down these arguments — not just in the US like they started to this week, but in the EU where the laws are more stringent and regulators are less sympathetic to big tech — what are the options?
In my view, regulators won’t be able to solve it directly. Even if they do get a win, it will probably mirror the laughable $5 billion fine that Facebook got a year ago. Yes, $5 billion is a lot of money, but for Facebook, an $800 billion company, that’s a small toll on the road to world domination. I’m sure if Facebook could prepay $5 billion for their next violation they would — but I digress.
History would tell us that the most effective way to disrupt incumbents is to compete in adjacent categories, establish a wedge, and then expand. The catalyst for this is usually the emergence of a new technology that enables new consumer behaviour that is a 10x improvement. If you try to compete with incumbents using all the same tools that is a losing battle because they have already mastered the tools and they have the audience. A good example of a disruptor is one of our present day monopolies: Amazon. The emergence of the web browser and efficient supply chains enabled Amazon to offer a store-front in everyone’s home and fulfill those orders relatively efficiently (by ’99 standards).
Existing book stores, and later department stores, tried to just copy/paste their business into an online format and failed. Amazon didn’t have the organizational baggage — internally and externally — that makes it difficult to turn a big ship. They were nimble and able to navigate the emergence of e-commerce like a speedboat. Their competitors had legacy supply chain contracts to navigate, a team not rostered for e-commerce, a customer base that had expectations, and lastly, didn’t have an incentive to disrupt themselves. Classic Innovator’s dilemma.
Amazon coming to market brought innovation through their customer obsession. They offered a net better user experience — cheaper products accessible anywhere. They were disruptors and celebrated for it. However, as Harvey Dent presciently said: “You either die a hero or you live long enough to see yourself become the villain.”
Amazon went from disruptor to monopoly. Now they are using their position to suck the air out of the room from anyone who would dare challenge them.
Example 2: Take seller data, build competing products, usurp distribution position on Amazon.com
Shady, yes, but the argument goes: we are customer obsessed, and we are delivering better, cheaper products to consumers because of these actions. Hard to argue that their customers are net-worse because of it. But again, what’s the deadweight loss of those fledgling startups that were lured into the cavernous Amazon HQ and asked to disrobe for “due diligence”? We might never know. Maybe they would have gone on to build incredible products that changed how things are done in that category, maybe not — but one thing is for sure, Amazon is looking to add more cogs to their machine. They aren’t looking to build a new machine.
In my view, yes, these practices by monopolies may stifle innovation, but we should not blame Amazon, Google, Facebook, etc. They are acting rationally. They see the market opportunity and they are doing everything they can to deliver the very best experience they can to consumers. But the best experience they can deliver isn’t necessarily the best consumer experience full stop. It’s misaligned incentives, and we shouldn’t expect to get to aligned incentives. Running a profitable business is not an altruistic pursuit, nor should we try to make it that. The best products come from intense competition.
To drive better competition, the question we should be asking isn’t: how can we slow monopolies down? We should be asking: how can we speed everyone else up?
This is one of the reasons I get excited about crypto. The headlines in crypto are largely driven by the price of crypto assets. This is a symptom of a new feature set that money never had before. It is open, it is global, and it is always on. These three points speak to the access created by crypto. Starting anything new is challenging, and when the total addressable market is constrained by something simple like payment rails, that becomes increasingly difficult. Many software businesses can serve the needs of global consumers, but their ability to monetize those consumers makes it challenging.
Access is the first piece. The second, and possibly the most impactful, is the alignment of incentives. If everyone in a network programmatically shares in the value created by the network that creates a strong incentive to not only participate in the network, but actively try to grow it. Imagine to advertise through Google search an advertiser had to pay in a token. And then imagine every one of those tokens was distributed to the users of Google search. And imagine there was a scarce amount of those tokens, so as demand for advertising tokens grew, the value of those tokens grew. The early users of Google Search who had some of these tokens would be sharing in that upside, and they would be incentivized to get their friends to use Google Search because that would attract more advertisers, which increases the value of the ad tokens. It would create a virtuous cycle of people bringing more people to the network and capturing commensurate value for that.
The example above is just one of many ways that crypto could align the incentives of the network participants. Jesse Walden articulated this well in his post The Ownership Economy: Crypto & The Next Frontier of Consumer Software
“The ownership economy doesn’t always mean a literal distribution of tokens, stock options, or equity. It also doesn’t necessarily mean that an application or service is entirely built on a blockchain. Rather, it means that ownership — which may manifest in the form of novel economic rewards, platform governance, or new forms of social capital — can be a new keystone of user experiences, with plenty of design space to explore.”
As Jesse noted, there is plenty of design space to explore. We are just scratching the surface of what’s possible, but the rate of innovation has been promising.
Ok, so why is this relevant in a blog post that started with antitrust actions against Google? It’s relevant because it shows a plausible path to a competitive marketplace that doesn’t rely on throwing sand on the road in front of monopolies. What makes this interesting is that while this business model is plausible incumbents would never adopt it. While the 2017 trope might have been: “put the business on the blockchain”, that’s similar to the ’99 trope of “putting the business on the internet”, just way more drastic.
Existing businesses have shareholders with a right to future cash flows. If all cash flows were pushed into the network of users they would be effectively taking their marketcap and dividending it out to the users of their product. Only a few companies would be crazy enough to disrupt themselves. Again, classic Innovator’s Dilemma.
For this to work it needs to be built from first principles. What is exciting to me is that the tools to do so are here today, and the networks are starting to emerge.
Antitrust isn’t going to deliver more value to consumers. As Google presciently stated in their response to the suit:
“This lawsuit would do nothing to help consumers. To the contrary, it would artificially prop up lower-quality search alternatives, raise phone prices, and make it harder for people to get the search services they want to use”
I agree. Propping up lower-quality competitors is not what we should want. And slowing leaders down is’t what we should want either. Let’s accelerate.
