Open Up Vs Break Up
There have been many calls to break up the large Internet monopolies; Amazon, Google, Facebook, Apple, etc.
Breaking up a large monopoly feels like a very 19th/20th century move to me.
I would prefer that politicians and policy makers think about opening up as the better intervention.
A good way to explain this is to go back to the architecture that Twitter used in its early days when there were many third-party Twitter clients. Imagine if Facebook, Instagram, Twitter, LinkedIn, etc were protocols, not applications, and there were many high-quality clients to participate in these networks.
Then the clients could innovate on things like content filtering, promotion of high quality content, business model, etc
If we are going to “break up” these large social media platforms, I would urge elected officials and regulators to think about pushing them to move from platforms to protocols instead of just ripping them apart.
We could do the same thing with search. Our portfolio company DuckDuckGo has built a nice search business by building a different user interface on top of one of the two leading search indexes. If we made it easier and reliable for others to innovate on top of the core search engine, then there might be many more options in search.
In mobile, a good first step is to open up the app stores and allow the browsers to have the same access to the operating system as native mobile apps.
In commerce, if I could checkout as easily everywhere as easily as I can on Amazon, there would be more competition for my shopping dollars.
I think you get the idea. It is very true that the big Internet services have built centralized monopolies and have consolidated their market positions. We do need more competition in these core services. And the best way to do that is force them to open up their services, not break them up.
But wouldn’t that affect their monetization strategies?
What about customers of B2B SaaS companies? They write a lot of customized software to integrate them into processes, and then if the SaaS changes policies you are out of luck.
Great point. On the same lines, an Indian guy in the UK (with an Oxford degree) contacted me some years ago, about his planned startup (HR / recruitment related, IIRC), asking if I was interested in being a co-founder or senior tech guy for it. I asked him about his business model. His startup’s plan was to piggyback on LinkedIn’s platform. I thought about it for a while and then declined, saying that the reason for declining was that LinkedIn could pull the rug from under him at any time by some change of policy, even if not deliberately designed to affect him. Later I read about some other startup(s) that failed for exactly that reason.
Exactly as though an evolutionary timeline died off and not only an entire species, but series of species based on ecosystems die off. Although that rarely happens overnight, the way it can in digital networked ecosystems.
Exactly! I can’t see a way to both force a company to provide an open API and not hit it in a major way. Being closed is one of their advantages which in turn secures their way of monetisation (ads, certain level of fees)
How would opening up something like the shopping cart be materially different than the existing amazon checkout app for 3rd party websites, or opening search be different from DDG using a bing API? There would be more players using the same APIs for front end adjustments, but assuming there are revenue models behind the use of these, it still feels pretty centralized (which I’m neither arguing for or against here).
I wonder why we would wait until specific companies get too big. Seems like the markets don’t like surprises, so maybe a less reactive and more proactive policy would make sense for the future. I have no idea what that would be, but it would relate to standards or access or market share. Now that I write that out, it terrifies me to have politicians making those calls. 🙂 I’m willing to have a monopoly or two as a price for free capital markets.
.Who is the scorekeeper on all of this?JLMwww.themusingsofthebigredca…
Having politicians making calls on technology that they cannot, and will not understand is why there was the great Microsoft breakup attempt, the continued battle for Net Neutrality etc. Tech is no more their space than logistics or entertainment. The government could do well to concentrate on components that they can understand and legislate. AT&T was broken to bits because of its monopoly status. What happened to innovation? Without funding for Bell Labs, all the innovation went to the fledgling tech companies. Perhaps it is time for the next wave of innovation, perhaps not. I for one would prefer that Apple, Amazon, Google etc get better at tech and find a way to partner with governments to identify bad actors (scammers, pornographers etc) than to waste resources and fall by the wayside like AT&T.
It’s called equilibrism. We need a rethink in our approach to networked ecosystems; particularly as the downsides to the digital ones are quite large. Enjoy: http://bit.ly/2iLAHlG
Amazon Pay provides checkout for 3rd party websites, using an Amazon account holder’s credentials: https://pay.amazon.com/us
But, Amazon won’t give me everything they know about me so I could in turn give it to Jet.com and they could recommend stuff for me.
I hadn’t thought about that, or whether or not that’s a good idea, but here’s how to access and download your Amazon order history: https://www.usatoday.com/st…
That is what you do when you aren’t the market leader. But that is what you should continue to do.
https://simplecast.com/s/80… podcast on breaking them up I linked to yesterday. I as at the Stigler Center conference on Data and Antitrust. It was very good and presented both sides of the argument. I walked into the conference thinking nature should take it’s course with the creative destruction of companies; and walked out thinking that some baby steps might be used to spur that competition. Portability of data is a big one.
Thanks for these references. All socio-economic and political institutions and frameworks are dominated by network theory. And networks always tend towards centralization (monopoly) without any equilibrating forces; at least the ones we see occurring naturally.
The media model is a snake that eats its own tail.And none of these companies know how to do anything else from a position of control.ButPerhaps the lever of open up or break up will force the right behavior for different reasons.Dunno who if any of the lawmakers can even articulate this though to make this a reality and whether any of these companies have the guts to do this proactively.
Thanks for this, Fred – I’ve explored the concept of a ‘Reverse ICO’ before which expands on the ‘tokenising the enterprise’ concept driven by Trent McConaghy at Ocean Protocol, and your thinking behind Steem and Online Publishing (https://avc.com/2017/01/onl….Reverse ICOs Part 1: https://medium.com/@harrymc… /reverse-icos-the-soundcloud-exit-strategy-almost-no-one-is-talking-about-651f4ef1f88eReverse ICOs Part 2: https://medium.com/outlier-…Reverse ICOs Part 3: https://medium.com/@harrymc… /reverse-icos-part-3-db72547afd87I’d be interested to hear views on this approach from you and the AVC community!
It seems reasonable and there’s precedent for it. They’ve done it with telephone companies, forcing them to open up their protocols. And television networks also were forced to open up. It’s worth a shot. But they should also be forbidden to do other things: i.e. Google Travel, Google Flights, etc.
A DOJ requirement to divest assets is one extreme, while a requirement to spinoff assets and have them operate autonomously could unlock even greater value for a parent company. What’s App, Instagram, You Tube, etc., trading as separate stocks could spread market share and in theory create more competition, while not diluting and possibly increasing overall corp value. A not likely scenario, but perhaps a possible one.
The AT&T divestiture was a horizontal spinoff of assets and a vertical disintermediation on two fronts: a) forced wholesaling of AT&T WAN or IXC network, and b) dial-1 equal access. So 3 moving pieces.There is still mandated interconnection; but hardly enforced. And the Republican FCC killed off equal access circa 2002-2005.Thankfully, we created plenty of layer 1 competition before we did that and the growth in mobility and internet ensured reasonably robust facilities based competition in the WAN.And market forces along with some unintended govt regulation (Part 15, which enabled WiFi) allowed Steve Jobs to single-handedly resurrect equal access in 2007. With that, the mobile oligopolies/monopolies were eviscerated above layer 2 as OTT stole the show.But it is safe to say that absolutely NO ONE foresaw in 1984, or even 2006, what 2019 would look like from a telecoms or network perspective.So the best way to think about “regulating” the internet monopolies is to open up the edge monopolies to interconnection and resale, while mandating that these core monopolies (and others) pay for access to these networks. We need settlements both east-west (from core to edge, or edge to edge) and north-south (from app to infrastructure and vice versa), in order to do away with monopoly bottlenecks anywhere in the informational stack.
This open up approach aligns very much with how they call their main products platforms. A platform has endpoints for consumption and/or interactivity.
I wonder how much annually these monopolies collectively spend on lobbying gov against the populist temptation to break them up? Huge sums i assume. Huge.I would first favour opening them up, see how it goes, and then consider alternative approaches if the desired outcome is not seen. I think it’s critical to have a very clear idea of what the desired outcome would be before taking action. That would require a 20/ 20 view of what the problems are, but the monopolies are working hard to hide the problems from view. Google tries to present itself as not being a search dominant corporation (the Alphabet deception). Facebook is now a very big book and definitely needs ‘editing’.
Per the discussion last week on regulating crypto… While the idea of carving up an enterprise along functional lines seems intelligent, who in the government is going to dissect these business at that level and then make a “call” that is fair? I just don’t see it. I would love to see FB and Google have to compete based on features and functions.When Standard Oil was broken up John. Rockefeller just got richer. I think Zuck and Jeff would survive just fine.
This feels like the tech version of forced regime change.
Fred, based on your impression from Mark Zuckerberg/Sundar Puchai Congressional testimony, what changes do you think would need to happen for something like this to happen?And is this post suggesting that companies like Facebook/Amazon are natural monopolies like my local electricity company?
I’m not even sure what “breaking up” something like Twitter or Facebook even means. It’s just a giant pool of people on a monolithic platform (mostly shouting at each other). Are we going split the Republicans from the Democrats? The Americans from everyone else? What does this breaking up even mean? Splitting up Facebook’s ad revenue from the Feed? Then you have a company that doesn’t work. Is there a functioning company left when that’s done? Splitting off Instagram and Oculus from Facebook changes nothing, it’s just nibbling at the edges.I’m not sure that opening up Twitter (again!) to 3rd party clients is really going to change anything. Even if there’s 50 “Facebook” clients, all the revenue (and hence all the bad actors) will still be chasing that platform/protocol. The ubiquity is what creates the value, and hence the problems. But in a world where 60% of articles shared haven’t even been read first, arguing about platforms and protocols seems to miss the point.The real change will come from people understanding that too much social media is bad for them. Social media can be a force for connection as well as divisiveness. One cigar a year isn’t going to hurt you much, but two packs of cigarettes a day will kill you.We didn’t affect smoking rates by outlawing cigarettes (mostly), we did it with education. Something similar is needed – creating a real recognition that social media in fact can be very bad for you, and that much of what (and who) is there is simply false and/or agenda-driven.
Good point. Generally there are 2 ways: horizontally or vertically. But you point out a 3rd way, diagonally, or across market segments (Z axis) in a 3D network model!
🙂 when I said I’m not sure what it means, I was actually serious 🙂 What would a horizontal or vertical breakup of Twitter mean? and would what is left actually be a functional twitter? What do advocates of breaking it up mean?
Somewhere here PointsnFigures references 2 sources where you will find your answer. My own view is that the breakup will actually come from somewhere else. If you read my archives on my site, I spoke about breaking up the WinTel monopoly by breaking up the last mile monopoly back in 1997. It actually occurred in 2007 with the iPhone. I reference that and the AT&T breakup elsewhere in this thread to Saltshaker.
The problem today is that Facebook and Twitter are anti-social networks, who have no imagination as to how the use ML and AI to improve people’s lives.LinkedIn and Instagram are better because the are closer to true communities, but they too need to deploy ML and AI in a way that enriched the experience.#its the engineers stupid.
.In taking legal action against a corporation, one has to start at the end game and reason backward.What future condition are we looking for that would be better than what we have today? If changes were made what would the “post-change” environment look like.This change is — has to be, can only be — justified by anti-competitive and monopoly pricing considerations, so declaring something a monopoly has to be supported by evidence.What is the evidence?In competition in the marketplace, there are winners and losers — that is capitalism. Sometimes, the winner “takes all” meaning some effort is so superior to others that it “shoots the moon” and takes all the tricks (for you bridge players).I suspect this is even more true when a bit of tech is mixed in. I use as my exemplar the spread sheet software business. I started out with VisiCalc went to Lotus 1-2-3, and ended up with Microsoft Excel.Sure, there are some new spreadsheet products creeping up, but Excel — for me — has become like a language, like using English. I don’t either need or want a new language or a new spreadsheet.At every step of the evolution, the capabilities got better until I could, literally, not use anything beyond pivot tables.Now, MS sends Excel out bundled in MS Office that has other bits of software that I use and others that I never open. I have to buy the package, but the price is very competitive.There is only one way I can buy it and only one entity from which I can buy it and every so often they force me to buy the newest version by stopping support for the older version — clearly anti-competitive control, but here’s the BIG QUESTION — Am I damaged in any way? No.In just the same way that some want to regulate crypto in a different manner than, say, securities or currency — take your pick — I think we are at a similar crossroads as it relates to the capabilities of these large companies.Ironic? — this blog calls for the reduction of the regulation of its own oxen while putting the yoke of regulation on those of others, proving that some things never change. Slightly worse, it calls for the use of an “outmoded” regulatory tool kit like anti-trust while also being a free range proponent of capitalism and innovation. Odd bedfellows. Of course, I would do the exact same thing.Other than their “offensive” largeness, these companies don’t seem to be exhibiting traditional anti-trust or monopolistic behavior and their pricing — much of what they provide is lay down FREE — doesn’t seem to be onerous given the product and results.One thing is for certain — this will be a political football until the politicians figure out the money interest. The big tech companies already have.I think we are still in the “foot fault” range of transgressions.JLMwww.themusingsofthebigredca…
In other words, Facebook is a monopoly today as much as Seinfeld was a monopoly to Thursday Evenings in 1995.
Maybe there are some things that Amazon or Google COULD do that might concern the DoJ: I DOUBT that those things are being done or will be done, but, say, Amazon might tell a supplier that supplies, say, 80% of their output to Amazon that they (i) can’t sell direct and (ii) can’t sell to anyone else, or at least not for prices as low as they are charging Amazon. I DOUBT that Amazon is doing that or WOULD do that. But if they did such a thing, then that might be a case of “restraint of trade” or some such.Or that Amazon is big, successful, sells 500 K different items, often or usually has nicely low prices, what’s to object to, that is, why is that a bad sense monopoly?All this anti-trust against “big tech” sounds like the usual: Some politicians want publicity and, to this end, have a story they want told, a story about big, bad tech and the newsies very much want a story to tell. Sooooo, presto, bingo, we get stories, lots of stories. As such things go, in a few weeks the stories will die out and we’ll get a new crop of stories.Or the newsies write stories everyday, that is, everyday have pages and minutes to fill. Hopefully they write the best stories they can. Then, when the stories are really trivial, as in the monopoly threat of big tech, we know that every other possible story is less bad than that trivial, nothing story. Soooo, the newsies are confirming that the world is in terrific shape — that is, the worst story is big tech which is trivial.Of course there really are important, non-trivial, stories to tell, but those are well above the grade level of nearly all the newsies.
First off, winner takes all is imbued in ALL of humanity’s socio-economic and political institutions and frameworks. There is no equilibration process that we find in nature to counter the imbalance between the geometric value captured at the core and top and the more or less linear cost borne at the edge and bottom of the network. The result is hollowing out of the pareto optimum and skewing of normal distributions. The end-game is collapse and rebalance at some point. Humanity is doomed to repeat this cycle; only this time it will be much, much worse.Second, “I started out with VisiCalc went to Lotus 1-2-3, and ended up with Microsoft Excel,” basically says that you, along with just about everyone else and myself, had little say in the matter. You were pulled into Excel by the same network effects process cited above. This is from someone who “peaked” in 123 back in 1983. 🙁
I don’t see monopolies here, but I do want to help competition. I think Fred has regulatory oversight wrong.Time is the equalizer. Given enough time, many of the core elements that make these big tech companies desirable will fall away.You won’t need data centers to index the web for everyone. The social network for everyone, will be less attractive than a social network for no one. Smart phones will be commodities. These trends are the result of the churning machine we call modern society.
I agree with Fred that forced mitosis of large companies seems like an myopic, outdated move. (What did we expect?) It seems to me that antitrust has been historically used to plug in those holes in capitalism where companies tend toward monopoly.However, today, we can use technology like decentralized protocols (markets) in order to break down monopolies and create competition in ways that we haven’t been able to do before. Will we take this opportunity?
Giving browsers the same access to the operating system than native apps, is a security risk. Probably that is the reason why they have not done that. The app store (at least Apple’s) is almost a warranty that people are not going to be victims of malicious applications.
Fred, how exactly does the government force a company to “open?”What I’ve yet to see anyone write about is the idea of horizontally breaking up the big software-based companies. Don’t split Facebook into Facebook.com, Instagram, and WhatsApp, but instead, like Standard Oil, split it up into five (or more) pieces, each with a copy of the source code, each with a fifth ownership in the copyrights and patents.The results of that are easier to think through for Twitter. You wake up Breakup Day 1 and your Twitter feed is now on Twitter One, mine is on Twitter Two, some on Twitter Three, some on Twitter Five. The Twitterverse split in shards. What would the CEO of Twitter N do months before?She’d sit in a room with the other Twitter CEOs and have hashed out an API to let Twitter A users follow tweets from Twitter B. But more, one of these CEOs would have met with Zuck and done a deal to put their feed and followers into Instagram. One would have met with Google to get their feed back into Google search results. Etc.If perchance Facebook were similarly split into M Facebooklets and Google split into five 10^100/5 googles, we’d once again have a thriving set of competitors trying new combinations and offering open APIs to attract new ideas.The only downside would be a few years of messiness as we lost track of a few follows and followers. In terms of economics, Standard Oil gained 5x in value three years after its breakup. AT&T gained more than that. Competition begets innovation. Innovation begets market value.Winner takes all doesn’t mean the winner provides the best possible product, only that we’re unlikely to ever see another solution provided.