- You can make a profit in the industry of your advertisers. This is a huge if, and is impacted by scale, expertise, management focus, time/maturity, branding and many other factors.
- The return on investment you can make in industry X is higher than the return on investment you can make by expanding your operations in your current industry. This could be true for social sites, where increased investment (i.e. programmer time) might not get them many more users, but they still have to compare the possible revenue to the increased profit from firing some programmers and going into maintenance mode.
These to are the main reasons why you see companies outsourcing non-core activities (like cleaning and even product development), the markup they now have to pay on the service is lower than their internal cost of capital.
-The SN network in this situation would be acting as a "holding" company. It does not expand into anything itself. The child companies do, and gain expertise (by possibly hiring from competitors), and also advantage by more traffic sent to them.
-That return on investment would be higher, in the long run. Yes, thats an if, but its a much better if than "social graphs completely turning around CPM for something like FB".
Those are also, by their opposite, reasons why companies consolidate and acquire...
Your first point is essentially meaningless. It makes no difference if it is a division, a child company, or part of the core business. Someone has to think about it, manage it and execute it. All this costs money which you could use elsewhere _and_ it creates potential conflicts of interests with your own business. This means that you need a really good reason to go ahead.
For an example of conflicts of interest look at the Windows/Office vertical integration and the constraints placed on Office in order to keep Windows profits high.
Your second point boils down to the fact that social networks give an awful return on investment. Could be true, but if so that's more of an argument to get out of social networks altogether and find a more profitable business.
Microsoft is still doing well off of office, and acquisitions/ventures such as hotmail and msn/windows live search. When most "average" people use their computers for the first time, and then on, they'll use internet explorer which goes to the msn home page and searches on Microsoft's search solution. The "average" user out there still uses IE, not firefox or opera. As a result they get a ton of traffic they wouldn't otherwise get.
As you can see, Windows Live is number 6 on there. It is terrible compared to google, but that is still alot of traffic they get from synergy between their companies. Would Microsoft make more money if they instead sold their search traffic/IE users to google? Not at all.
This is very silly. It's a well-known concept in the business world that companies should stick with their core competencies. Ones that try to do too many disparate things at the same time generally fail. Facebook's core competency is not selling mortgages or credit cards.
Also, if mortgage brokers are only willing to pay some portion of 14 cents CPM there, that means Facebook is creating very little value for them. Which also means that if Facebook got into the mortgage selling business, they could probably capitalize off of it in much better ways than just advertising it on their relatively worthless site.
So in essence, the only industries Facebook would benefit from entering are ones that already pay them a high CPM. But if people were paying them a high CPM, they wouldn't be in this predicament.
"Facebook" wouldn't have to change their core competencies at all. It would have a stake in a company who's core competency is whatever that industry happens to be.
The site is indeed relatively worthless, and I doubt anybody at all is paying them a high cpm in anything. Thats the whole point, to turn low cpm into somewhat higher cpm, in the long term by capturing more of the revenue stream.
Is it fool proof? No way. Facebook or whatever, could risk losing their investment in whatever company they bring up, and also lose the opportunity cost on whatever ads they could be selling instead of sending to their own child company instead... But does it stand a better chance at making more money in the long run compared to magical BS like social graphs and social ads? Yes, I definately think so.
The minute a startup starts making significant investments in mortgage brokers, credit card providers, etc., you're leaving their core competency.
I don't think you're understanding why this provides no value to anyone. You're turning Facebook into a Berkshire Hathaway of sorts, with the only coherency being that every company they buy a stake in advertises on Facebook. By that logic, why shouldn't Time Warner start buying stakes in whoever advertises on CNN? Why shouldn't everyone who sells ads do this?
Facebook gains no higher CPM from doing this either. I mean, they may make more money if these auxiliary businesses do well. But it's no more than if they kept selling ads on Facebook for 14 cents cpm, set up these side companies, and spent 14 cents cpm advertising them somewhere else. The fact that they own Facebook provides no value whatsoever for the side companies.
Effectively you're saying that any company with a non-functioning business model should just do something else too. That doesn't solve the problem.
However, I'm trying to make this "whole is greater than the sum of its parts" argument. But, with Berkshire, almost all of its investments thrive well on their own. Buffett invests in each and every one because of fundamental reasons within each individual company.
My argument is that some companies are making a lot more money off of each traffic batch than the traffic batch is actually worth. By figuring out such companies, the value of the traffic to facebook would then be higher. Would this work with every company that advertises on FB? No. Could it might possibly maybe work with at least one or two companies? Yes.
Saying any company with a non-functioning business model should just do something else is really stretching it. I believe FB does have a functioning business model that could use a great deal of improvement. This, to me, could be such an improvement. What other alternatives are there? You have already said yourself that things like "social graphs" are just BS meant to inflate valuations to $15 billion.
I agree that social networks need better business strategies in general, but if the advertisers are making more money than the social networking sites, why make the SN site in the first place? Why not just try to beak your way into porno, gambling, credit repair (the list goes on) business and just sell adds like everyone else?
I think the answer is that we believe SN sites should be able to make more money than these other businesses. The person who figures out how to make that happen will be rich, but I don't think the answer is to just add another business onto your SN.
Sell ads on what? Those sites are buyers of ads. They do the conversion of traffic.
I should have made more clear that this is a longer term strategy, and yes, indeed more expertise is needed in these industries. However, what these industries do is not rocket science, and can be reproduced over time.
My point is that SN sites could eventually do the conversion of that sort of traffic themselves, and gain higher CPM In the Long Run. This process would take years, but is rewarding in the end.
But it's not really a higher CPM at all. There's no benefit to Facebook starting a mortgage brokerage and advertising that on Facebook (which apparently isn't a very valuable seller of advertising, or they'd already have decent CPMs) over just switching to a mortgage company entirely and advertising on other social networks. Or better yet, switch to selling mortgages and advertise on sites that are valuable to mortgage brokers.
You're assuming perfectly priced traffic, in an arbitrage-free environment where the buyers of traffic pay perfectly for the traffic, an exact amount proportional to their revenue that they will get from that traffic.
It doesn't always work that way in the real world. Some random company buying the ads at 10 cents CPM is not always getting traffic that is worth that much 100% of the time. By actually getting the entire revenue stream, the "child" company, and thus its parent would capture the true value of the traffic over a long period of time.
No, I'm assuming that they could just buy ads with the same ROI in many places. For instance, if Facebook ads cost 20 cents cpm, but they earn 40 cents, then their ROI is 2x. I would assume they would have no trouble finding ads elsewhere with a 2x ROI. In fact, I'd be surprised if the market didn't function to make their ads be a 2x ROI at MySpace, Bebo, Hi5, and maybe Digg, Reddit, etc.
Spotting market inefficiencies like that should be really hard to do, considering that all of the mortgage brokers, for instance, already know their ROI on various sites, and the market will make it roughly equal everywhere over time.
Your putting too much faith in the buyers of ads reacting to the changes in prices and knowing exactly how much the traffic they are buying is actually worth. All mortgage broker's/free i-pod scam/dating site/etc know the exact ROI they get back from each and every little bit of traffic? No way. In some industries, theres even a bunch of buyers, some less sophisticated than others, but are yet still able to buy the traffic at a certain rate because they are making a profit at that rate. That isn't optimization. In industries with fewer buyers (high traffic dating sites), one site might have to purchase traffic from a multitude of different sources, or indirectly purchasing it from their own affiliates who might purchase it themselves, to keep up their market share against other big competitors.
This isn't about spotting temporary market inefficiencies, its about spotting long term ones that lurk beneath the surface.
In terms of real money, what comes out of it, is affiliate sales for companies + brand advertising.
Brand advertising is mostly worthless and gets talked up in value because it is even more worthless on national television.
But the kind of advertising that results in a sale is worth real money. Facebook either needs to start selling mortgages in house or develop better relationships with someone who sells mortgages.
Yeah, I haven't even added in the thought of brand advertising as well. That would add to the effect, as the child company also gains eventual brand recognition over time in that situation.
Interestingly, parallel programming theory comes into play here. There is a sweet spot between completely distributing an algorithm and completely consolidating an algorithm that allows it to scale effectively with more processors.
I have recently had the same idea and tried to execute it with as an astonishingly great failure. I have a real estate site with high traffic, I sell adverts and have adsense too. I suspected I could see directly to these visitors with another spin off that I controlled. So I built an e commerce site selling organic wine (which I suspected would be of interest to the same clientel) and advertised it prominently on the real estate site. Hardly any traffic has resulted from the advert. The bottom line is, if the advert is not on topic with the mode de operation of the user at that specific time it will be ignored.
He has two assumptions:
- You can make a profit in the industry of your advertisers. This is a huge if, and is impacted by scale, expertise, management focus, time/maturity, branding and many other factors.
- The return on investment you can make in industry X is higher than the return on investment you can make by expanding your operations in your current industry. This could be true for social sites, where increased investment (i.e. programmer time) might not get them many more users, but they still have to compare the possible revenue to the increased profit from firing some programmers and going into maintenance mode.
These to are the main reasons why you see companies outsourcing non-core activities (like cleaning and even product development), the markup they now have to pay on the service is lower than their internal cost of capital.