A Brief List of Business Units Google Could Be Separated Into
A Brief List of Business Units Google Could Be Separated Into
For no particular reason.
Well, Google may be under some sort of Justice Department investigation.
I’m pretty sure this will end up with Google assessed some paltry sum that they will be able to pay off with the amount their stock goes up once investors see there are no consequences to becoming Lexcorp. (I’m going to refer to ‘Alphabet’ as ‘Google’ throughout because Alphabet was a tactic from day one.)
But I do love me some fantasy antitrust case adjudications, so let us think on this for a hot minute: What would Google look like if the United States decided to break it up? What units would Google/Alphabet get broken up in to? This is a fun game, but good games need rules:
We’re being fair: This isn’t just an excuse to give Google the finger. Each business unit has to be at least as functional as a Baby Bell and have the potential to be a stand-alone business.
This is my theoretical fantasy land where we have a working government and judicial department, so we’re not going to worry about them merging back together like the Baby Bells did. Also, this would inevitably be followed by antitrust examinations of the rest of Big Tech and likely some serious changes to those companies.
Each unit must be logical, that is representing a whole component of business operations that can operate on its own, and also have a self-contained set of purposes and goals that, when independent, have incentive to act competitively in the marketplace.
The breakup must serve the American Consumer. The broken-up company should end up with a set of organizations that make the marketplace more competitive by opening up spaces for new companies. The breakup must also create opportunities for the consumer’s buck to go further. Finally, it should also make the consumer safer and more private.
OK, with these rules in mind, off we go!
Since we’re going to be American Superheroes here, this music break is brought to you by the most American, modern Superhero who watches over Wall Street: Your friendly neighborhood Spiderman.
Drive - The Professional Services Unit
This is an easy one. The most logical stand-alone service provider is the one that has basically been languishing under Google for years now. Drive does not seem to really benefit from developments in other units and would be the core of the PSU (Dropbox proves this is a fully functional stand-alone business). It would be joined by Google’s entire tool-set of Docs, Sheets, online programming tools, like Google Colaboratory, Hangouts, whatever other chat/conference applications are still active and, most notably, Gmail and Calendar.
Gmail has basically languished for years. It has lacked competition and so its capabilities are slow to develop. Google has little incentive to do anything but make it better for targeting and showing ads. Pulling the Gmail system out of there is needed to assure this breakup enhances user privacy. Housing it with the Professional Services Unit makes the most sense. Gmail is basically the loss-leader for Google Drive and always has been. Good fit there. Hopefully we’d get a better Gmail product and create a competitor for that space that should keep prices lower.
The Gmail ad space will open to any exchange willing to meet its specs.
This will also be a good location for Google Fi as a useful rounding out of the services that this business could provide to both individuals and other businesses.
NewsTube - The Media Unit
YouTube is obviously dangerously overpowered. It needs to be separated from search and ads, but that doesn’t solve some of its core problems. It needs some synergy from another part of Google to make the platform actually make real money. The easy answer is that ChromeCast production could fall into this unit. But it isn’t just that. Google, along with the rest of the web, has been preaching the gospel of news video. And its major partners in any ChromeCast operation will include cable media.
YouTube’s success as an independent operation will come out of good, effective and ethical aggregation. You could say it has made a business out of hosting video, and it is that, but that isn’t really their business; anyone can host terabytes of video now. The goal of YouTube as a business is to keep you engaged by stringing together different content sources effectively. The longer you watch, the more ad dollars. That aggregation includes handling content from news organizations, so it makes sense to pair YouTube with Google News. Google News is also fundamentally an aggregation unit that can make a profit via interstitial advertising.
The NewsTube unit will be formed around the theory that multi-platform, multi-format aggregation of news and related media is an operational business. Others seem to have pulled it off, though they hit up against the wall of Google, so this will create brand new energy in the marketplace. Competitive pressure on this unit will force it to reconsider how it understands itself as a host vs an editorial organization and require it to differentiate itself via moderation and selection, instead of being protected from competitive pressure by living relatively rent-free on Google’s servers.
It will require two major changes:
YouTube and Google News will have to open up their advertising systems to any effective service providers, forcing them to state technological requirements and collaborate with ad tech providers to move the ad tech space towards more effective video-focused technology overall.
The Search product will need to restructure how they handle news to be a more abstract user-controlled RSS Feed Reader-style product and open its architecture up to any organization who wishes to provide editorial-aggregation-as-product (EAaP) to that space. If Search wants their news product to be driven by Google News you can make that choice, but others can also engage in the aggregation space of the Search company, from Flipboard to the NYT. Aggregation becomes a competitive space within search and the Search unit will be forced to do clear rule-making that may give opportunities to other EAaP providers. Further, it opens opportunities for different types of aggregation products not deformed by Google News’s dominance in the marketplace.
Games and Coins - The App/Transactions Unit
The Google App & Money Skimming Business that is Google Play is clearly non-functional without transactional technology, including whatever Google Pay is languishing as and other payment and processing technologies. This becomes essentially a Steam-style service for apps and games. It would be easy to see them expanding their selection platform outside of Android in this situation, potentially providing a general Linux application storefront and eventually expanding to other operating systems.
Making this an independent unit that might end up having to compete with other operations (it is very easy to see Product Hunt become its own apps/games storefront in this world, as just one example) and would force quality control and security to the forefront as competitive concerns.
This is clearly an emerging business, Epic Games and Valve both see it as such, as does Amazon and, like Amazon, the addition of an overall transaction service provider will nicely bootstrap this into its own company while opening up Android for competition from other storefront providers who can compete on technological, editorial, or ideological differences.
The one inevitable hiccup here is people will claim that this opens up people to download more unsafe applications, as a centralized authority is thrown out the door, to which I will reply: How safe is the Play store right now? Not really that safe.
Search, Analytics, Finance, Translation, Maps and Servers - The Search and Engineering Services Unit
Search cannot work without innovation in servers because good search is expensive and requires a lot of CPU and storage. Analytics needs to be separated from the ad business for privacy and safety reasons; this is its most natural fit because it works on collecting data with the same logic that is used to determine the stature of pages’ rank in search. Translation is an obvious fit here; understanding webpages’ relevance requires translation tools to fill in the personalization gaps. Translation and Analytics also belong together because they both involve parsing webpages. Maps becomes a useless product without Search and Search would become significantly less likely to succeed as a business without Maps. Same can be said of Finance.
The Engineering Services Unit would also include all of Google’s web hosting tools, cloud computing for business and anything that is currently intended to compete with the AWS suite. Also, to whatever extent Go and other developer tools are still being developed internally, they would end up here.
The focus of this particular setup is, instead of dictating standards to others, this business is forced to collaborate and cooperate with the open web and browsers to create a more open taxonomy of the web; clearer community-supported engineering standards; and focused competitive products around engineering and web parsing. The goal with this unit is to diffuse how Google has used the leverage of Search in combination with Ads and Chrome to negatively impact technological standards while still leaving the new unit with enough capabilities to improve the web.
Search and the other products under this unit will continue to be a space for advertising, but now it must consider all technologically capable vendors for running that ad space. I suspect we’ll need to give them a year with the new Ad Services Unit running it so others can understand and spec up to the requirements and then require this group to open that space to competition.
Android & Hardware - The Phones, Chromebook, Nest, Google Home and VR Unit
I’m really tempted to put Android and the phone hardware in different units, but I don’t think Google’s phone hardware business is all that strong and I’d be wary of Android residing anywhere else.
Developing Android further as an operating system and the hardware into a better functioning business (as opposed to the irregular mess it is now, in part, because it lacks real market pressures) seems like logical paired opportunities and separating this business unit from the Play Store could potentially create a new set of Hardware/OS mobile devices, like we briefly thought was going to happen when Palm and FirefoxOS sort of looked like they might be a thing. All the rest of the hardware Google produces either pairs with their mobile devices or runs Android and, thus, should fall into this business unit. This organization could also realize some Red Hat-style B2B opportunities that Google isn’t required to bother with right now.
It is important to separate out the devices here that do data collection from the ad business for obvious reasons.
Chrome, AMP, Identity & Labs - The Chrome, X, Robot Cars and Other Silly Experiments aka Chrome Labs.
Chrome isn’t a business and AMP is somewhat community run right now. These aren’t really profitable components, so they best fit in the Bell Labs of this breakup, the experimental development group that’s going to land a lot of the weird Google patents that haven’t been realized anywhere. Licensing those patents and developing new car brains, space software, and whatever the hell else X Labs does would make a decent stand-alone unit.
Keeping Chrome the away from both ads and search is really important and AMP’s integration with search is a clear anti-competitive behavior right now, in that it wouldn’t work without control of Chrome, the Ad Networks that sell into it, the search page placement and the caching and server integration that backs it. Moving it to become a part of how Chrome talks about how they wish the web would work makes it a lot safer.
Stick Google Fiber in here because Google clearly doesn’t treat it like a business but much more like a weird experiment.
Wow, we’ve come a long way, and now we get to talk about ads! Let’s do a 2nd music break! How about some Latvian Bagpipes? Did you know that Latvia has its own type of bagpipe? They are called dūdas.
Also, a reminder that you might consider paying me for work like this:
So this is the ad tech section! This is important because I take a different view than some as to how Google’s ad business might be split up. Specifically, I don’t think it should all be its own thing. That would still leave a significantly anti-competitive operation. To really open up the ad tech marketplace to honest competition, while creating theoretically sustainable businesses, it needs to not just be separated from Google but also split in half. On one side, the ad serving technology and, on the other side, the ad exchange technology. This clears up some conflicts (like AdX having privileged status within Google Ad Manager) and has a logic to it; they are two sides with different requirements for success.
AdX - The Ad Exchange Unit
This unit would comprise all of Google’s current set of Ad bid and exchange products, including AdX, EDBA, GDN, DV360. This means all technology that manages bids, exchanges ads, understands user data, etc…
This is important because it separates the technology that transacts on data from the technology, like Google Analytics, which collects the data, mirroring the separation between those two sides in the rest of the ad tech space. It would send a clear message that 3rd-party data collection cannot be part of ad exchange businesses without causing conflicts of interest that harm consumers. It also removes a number of privileged anti-competitive system interactions, including: AdWords and Search, AdX and DFP, and a variety of Google-controlled demand matched with YouTube or AdMob. I’d also hope that it would push exchanges to be more competitive in a way they currently are not.
A while back someone I know suggested that all exchanges be non-profit and if we’re making up fantasy DoJ cases here I’d prefer that even more.
Double Click - The Ad Services Unit
This unit would be all ad services for other platforms, including what is now called Google Ad Manager (aka DFP): AdMob, Google Campaign Manager and AdWords. Running these systems within their own business will leave a lot of other products with empty spaces where they used to integrate Google ad tech. That would be awesome. Now those products have to examine a competitive marketplace for ad tools. They will be required to take general bids post-breakup from ad server providers, including this new business.
The ad services business still has enough successful contractual connections and revenue that it could continue to offer a free product. Removing its tight integrations with the exchange systems will relieve significant marketplace pressure that pushes people into Google ad products, creating an opportunity for others to offer competitive ad servers.
Google’s has failed to innovate in the ad services space (DFP has been almost entirely bereft of technological improvement for at least a decade) but retained their dominance in that space. This is a clear indicator of anti-competitive behavior by Google. Pushing the Ad Services Unit out on its own would require this new organization to push forward real technological improvements or (without the advantages of integration with servers, search or exchanges) face actual competition that might innovate on things like privacy and performance.
The Ad Units’ separation from Google and careful division is clearly the most difficult part of this proposal to consider. So much of what has made Google’s ad systems the most dominant in the marketplace is unclear, blackbox integrations and connections. We could look at this separation and say: Is even this enough?
If one company controls Supply and Demand systems, even without execution, that could allow them to place unfair requirements or control on the market. I think that’s an OK argument to make, but I think the ad tech marketplace requires real change from where it is now. The goal with the separation here is to create opportunities for these units to actually compete with other ad tech in ways that create market innovation and change.
The end result of all this should advantage general web users. Like I said at the top, these individual Baby-Bell-Style businesses should not be non-functional or without an opportunity for profitability. The separation I suggest creates two very different opportunities. One company would be forced to innovate and differentiate on their capability to provide service. The other would have to improve their technological performance within the larger ad tech landscape. A different split might be possible, but I think this is the most fair.
But I’m open to other ideas! Tweet them at me if you feel strongly that there’s a different, better split possible here. Also Tweet at me if you have any alternative ideas for the whole thing! I think we really need to talk about this in public because regulators, if they ever end up breaking up Google, will need a vigorous public discussion about how that would work.
This was very long, so I’m going to give you one last song. If you made it here, thanks for sticking with me!
Interested in talking more about this newsletter? Join the Keybase.io team to chat about it: https://keybase.io/team/gneist.newsletter