Fri

Dec 28
2007

Tim O'Reilly

Tim O'Reilly

Trading for their own account

I promised recently to publish more of my reflections on what I've learned from studying the parallels between Web 2.0 and financial markets, one of the real wake-up calls was the way that Wall Street firms moved from being brokers to being active players "trading for their own account." Ever since I heard Bill Janeway point out that over time, Wall Street "firms began to trade against their clients for their own account, such that now, the direct investment activities of a firm like Goldman Sachs dwarf their activities on behalf of outside customers," I thought, whither Google, Yahoo! and Amazon?

And sure enough, there is lots of evidence that this process is already far advanced. These sites, once devoted to distributing attention to others, are increasingly focused on consuming as much of the user attention as possible. What else do you make of Google's recent sally against Wikipedia, the so-called knol. Anil Dash's analysis was spot-on:

Google's announcement of Knol shows that they understand some of their key business drivers very well; With as much as 5% of the search result links for popular terms going to Wikipedia pages, a solution to capturing some of that traffic in an environment that Google can control and display ads on makes good business sense....

[But] Knol shares with Google Book Search the problem of being both indexed by Google and hosted by Google. This presents inherent conflicts in the ranking of content, as well as disincentives for content creators to control the environment in which their content is published. This necessarily disadvantages competing search engines, but more importantly eliminates the ability for content creators to innovate in the area of content presentation or enhancement.

Everyone applauds when Google goes after Microsoft's Office monopoly, seeing it simply as "turnabout's fair play," (and a distant underdog to boot), but when they start to go after web non-profits like Wikipedia, you see where the ineluctable logic leads. As Google's growth slows, as inevitably it will, it will need to consume more and more of the web ecosystem, trading against its former suppliers, rather than distributing attention to them. We already take for granted that common searches, such as for weather or stock prices, are satisfied directly on the search screen. Where does that process stop?

And much as I support what Google is doing with Google Book Search, I am troubled by the fact that they give preference to their own content repository over digital copies provided by publishers or other aggregators. (See my post Book Search Should Work Like Web Search.)

We see the same pattern at Amazon, which is aggressively pursuing authors for direct publishing on the kindle and seeking to displace publishers by making themselves the sole source for books on the device. [Update: And reader GL, in the comments on this post, observed "this year I found it quite interesting that YHOO made such a large effort to become a 'super affiliate'. The amount of product reviews tied to affiliate links was amazing. Would you rather make $2 per click selling PPC ad or make $25 per click (5% of a $500 sale by advertising along side other ppc advertisers? I think the answer is simple. Interestingly, in this scenario YHOO wins on both sides of the trade."]

Ultimately, I think we see this pattern in the economic development of every innovation. When a new technology is introduced, there's a lot of green-field opportunity, and so much value is being created that there's no need to capture it all. But as the technology matures, the winners need to capture more of the total value being created. They gradually crowd out suppliers as well as competitors.

Fortunately, there's another side to Bill Janeway's unorthodox history of Wall Street (as told in the Release 2.0 issue on the subject), and that's the one we're focusing on at the Money:Tech Conference in New York in February. And that is the rise of new opportunities for specialized information and data mining services that go deeper than what's available in search engines. I'll write about that next week.

tags: release 2.0, web 2.0  | comments: 30   | Sphere It
submit:

 
Previous  |  Next

0 TrackBacks

TrackBack URL for this entry: http://blogs.oreilly.com/cgi-bin/mt/mt-t.cgi/6171

Comments: 30

  Anil Dash [12.28.07 04:10 PM]

Thanks for the kind words, Tim. I think you amplified a point I was trying to make, but couldn't quite articulate -- this is pitting Google against some part of the audience it has tried to serve.

In technology, this has traditionally been about bundling; When Microsoft included disk compression in their operating system, or Apple bundled widgets, the primary victims of that bundling were the independent software developers. But by bundling source materials with the tools used to discover (Google) or consume (Kindle) them, it isn't just other search engines or ebooks that suffer; It's everyone trying to participate in the marketplace of ideas.

  Kin Lane [12.28.07 04:38 PM]

Isn't capitalism a beautiful thing, wait....is it?

  Joseph J. Esposito [12.28.07 05:49 PM]

The moral is, Google is the snake that eats its own tail. But all organizations that place themselves between the purveyors of original content and end-users have the potential of turning serpentine; the Internet, alas, has nothing to do with it. Thus in the hardcopy world of trade book publishing, the largest U.S. retailer, Barnes & Noble, is now also around the tenth largest trade publisher, competing with the very publishers that supply it.

This is part of the reason (not the only part) behind the AAP lawsuit against Google's book program and Viacom's suit against Google/YouTube. Whatever the legal virtues or limitations of these suits, part of the economic assessment is that today's cheery partners are possibly tomorrow's bitter rivals.

  Thomas Lord [12.28.07 06:07 PM]

There would be fewer accidental double posts in the comments if the web system used for the blog didn't have "sleep(60)" calls (seemingly) embedded in it.

-t

  Niraj J [12.29.07 12:39 AM]

What Google and Amazon are doing is a time tested business strategy - Vertical Integration.

Check out
http://www.gandalf-lab.com/blog/2007/11/google-energy-and-vertical-integration.html

Going by the Capital Markets analogy. The Full Service brokers have disappeared and the market is split into either transaction processors (discount brokers) OR Wealth Management experts.

Same fate is what Amazon probably expects for Publishers and Google expects of Content Distributers(like wikipedia)

Amazon and Google would probably like to take the place of Transaction processors.

  jen_chan, writer surefirewealth.com [12.29.07 02:05 AM]

Pattern in the economic development of every innovation will always be present. It is up to us to look at these trends and decide to teke them as they are or to try and clamor for something better than the prevailing pattern.

  Michael R. Bernstein [12.29.07 09:48 AM]

Tim, I'm not sure this is really tied to 'technological maturity' per se. For example, I see the same phenomenon in Barnes and Noble becoming a publisher:
http://en.wikipedia.org/wiki/Barnes_&_Noble#Publishing

  Tim O'Reilly [12.29.07 11:28 AM]

Michael --

B&N becoming a publisher is a sign of exactly the same phenomenon. They ate all the small bookstores to fuel their growth, and when that competition was gone, they had to start eating their suppliers.

But my main point isn't about at what stage this kind of behavior occurs, just that by looking at other sectors (like financial markets) we can see what is coming for Web 2.0 companies.

And like Jen, I wish that we could fight the trend by the choices we make, but I'm not sanguine. The only way that this kind of trend gets reversed is via a new technology giving us a "do over" (as the PC did with the mainframe market, and the internet did with the PC.) You get a period of innovation and openness for a while, till the winners start taking more value than they create.

  Michael R. Bernstein [12.29.07 11:57 AM]

I should have been more clear.

This is surely a result (and measure) of market dominance by a single company, but I don't think that market dominance is necessarily related to 'technological maturity' as you posited, (which to me suggests the usual model popularized by Geoffrey Moore).

  Tim O'Reilly [12.29.07 12:06 PM]

Michael,

You're right that it's not related to technological maturity per se, but to market dominance. The reason I put it in the context of technological maturity is that a lot of people, early in a new market, ascribe all kinds of idealistic new behavior to that market and to the technology that inspired it. But as the technology matures, all the old market patterns assert themselves in the new context.

  Ewan Gunn [12.29.07 12:30 PM]

I think it's unfair to say that centralising information is a bad thing, which is the implicaition in saying "We already take for granted that common searches, such as for weather or stock prices, are satisfied directly on the search screen. Where does that process stop?". I don't care where the information comes from, as long as it's authoritative. Imagine that Google 'bought' Wikipedia, and used Wikipedia instead of developing their own system - people would applaud it, it wouldn't affect them negatively.

There are two key things which I think have been glossed over in this: high-level authorship information, and searchability. The one fundamental problem with Wikipedia is its complete inability to search, something Google is very good at. A slightly lesser problem is that I don't know who 'wikimaster2007' is, but if I see an article written by the Professor of Particle Science at Edinburgh University - something I can verify, as an aside - I'm more likely to take it as wrote, further I can reference it properly.

There are also a couple of minor points that are very good: subject definition, which is excellent for searching concepts a user is unfamiliar with, which can easily be shown at the top part of the search results (with the option to hide); question and answer, again something very good if an article isn't quite clear on a particular point;

(Yes, I know there are problems with identity verification, but then how do I know Tim O'Reilly wrote this entry?)

The only potential downsides that I can see are that it doesn't seem colaborative - if I have a differing viewpoint I have to create an entire new article about it - and that there's no system to say which article is 'better' than another. I imagine that they will use the ratings system for this, but which is going to get the higher rating - the article written by the preeminent researcher in the field, or a laymans entry explaining it in plain language? Both have their merits, but there's no way to say which is 'better'. Finally, there is still the issue of privacy - once we navigate to a page that Google presents us as a search result, it can no longer track our movements from there (unless we return). However following similar articles, paths of interest through the knol system could quite easily be logged - in fact might need to be for individualised results.

As to the logic behind it, I agree 100% with you, but I think that is a major strength that isn't being played up by companies these days - imagine the Virgin Group amalgamating it's mailings to you when you are a customer in more than one product? Imagine also providing greater benefits when using more and more of their products? There are a number of groups that offer a multitude of services (Tesco is another that comes to mind), but none of them cash in on it quite like this. Hopefully with the coverage Google can give to this ideas someone in these groups will notice.

  Michael R. Bernstein [12.29.07 12:54 PM]

"But as the technology matures, all the old market patterns assert themselves in the new context."

I'll grant that this is the usual case, but is it inevitable? Cisco comes to mind as a dominant vendor in an even-more technologically mature market that, while few people would ascribe idealism to it, nevertheless I do not think would be described as a monopoly eating it's suppliers (instead, it is still eating new competitors) or 'trading for it's own account'.

BTW, the PC-to-internet transition you noted was *first* exemplified by walled-garden dial-up online service providers (eventually dominated by AOL), and only later transitioned to commodity internet (which predated it) and the Web. I'm not sure I would describe the second half as a technological change per se.

One thing that could create this same kind of not-really-technological phase-change in the future would be the resurgence of symmetrical broadband connections giving P2P technologies a boost (for example most web2.0 video sites are useless for distribution of legal high-quality video at a time when display sizes are only getting bigger), not to mention that just doing the initial upload to the cloud is a PITA when your upload bandwidth is constrained.

  j [12.29.07 03:20 PM]

Ewan Gunn wrote:

There are two key things which I think have been glossed over in this: high-level authorship information, and searchability.

My thoughts exactly. Primarily, the former is what I think the highest motivating factor for Google is and searchabllity follows right on its heals.

Outstanding, Ewan.

J

  GL [12.30.07 12:48 PM]

...."Wall Street firms moved from being brokers to being active players "trading for their own account."

There are couple reasons for this.

1) Market Makers - Firms such as GS create additional liquidity by becoming "market makers".
---- This is very similar to what Google, Yahoo and MSFT are doing. They are creating markets with incredible barriers to entry. Think NASDAQ, CBOE and NYSE versus Adwords, Publisher and AdCenter.

2) Traders - Trading desks utilize software, research and high "IQ's" to profit from "inefficiencies" in the marketplace. It just happens that a firm such as GS is probably the best as doing this.
---- GOOG, MSFT and YHOO also utilize the search created by their customers to find inefficiencies in the market to create products and compete directly with their customers. For example, this year I found it quite interesting that YHOO made such a large effort to become a "super affiliate". The amount of product reviews tied to affiliate links was amazing. Would you rather make $2 per click selling PPC ad or make $25 per click (5% of a $500 sale by advertising along side other ppc advertisers? I think the answer is simple. Interestingly, in this scenario YHOO wins on both sides of the trade.

But, the key here is to remember that the large portals have a oligopoly similar to the financial markets.

  Jesse Robbins [12.30.07 04:22 PM]

"When a new technology is introduced, there's a lot of green-field opportunity, and so much value is being created that there's no need to capture it all. But as the technology matures, the winners need to capture more of the total value being created. They gradually crowd out suppliers as well as competitors."

Similar to "Blue Ocean/Red Ocean".

  Ted [01.01.08 02:37 PM]

Tim. Sorry to be rude. Instead of blogging about tits and tats, why don't you tell your employees to write a better book.

Those web 2.0 theme books are awful. Especially the Ruby and Rails books. You also charged a lot for a thin book.

It comes to my attention that the quality is even worse than "For Dummies" series.

Please, please, please stop writing for a moment and back to Quality Control.

  mike p [01.01.08 05:16 PM]

Isn't pluging your Money:Tech conference the same thing?

  Tim O'Reilly [01.01.08 06:09 PM]

mike p -

I don't think so. When I plug an O'Reilly conference or a book, it doesn't take anything away from my customers. When a Wall Street bank trades against its customers, it's winning and they are losing. If Google no longer sends traffic to an external site, but instead delivers the data (and any accompanying ads) directly, they are winning and their former suppliers of destination pages are losing.

This really isn't about right or wrong though -- it's just about a market dynamic that happens. Understanding it, you can learn what to expect, and not be blindsided by it.

  Jerry Leichter [01.02.08 04:21 AM]

It's a mistake to look at this as a purely technological issue. Vertical integration isn't new. The economic drivers are obvious: Why share profits with suppliers who, in the end, won't provide me with exactly what I know is best for me anyway? Why not just do the job myself?

In fact, however, there are countervailing economic forces as well. "Knowing exactly what's best for me" tends to turn into "I know better than anyone else," which is rarely true in the long run. It also eliminates competition among suppliers - which inevitably raises costs in the long run.

The classic example of this is GM. GM grew both horizontally and vertically. I don't know if they still do this - haven't looked in years - but even 20 years ago, every GM car had a "Body by Fisher" plate on the door sill. Fisher was an independent maker of auto bodies; in the earlier years of car manufacturing, auto bodies were made by suppliers. Fisher was one of GM's largest acquisitions, and one of its last. It made GM the most vertically integrated supplier in history, measured by percentage of delivered value that GM itself had produced. That worked great for a while, but not that many years later, GM found that it had higher production costs than its competitors, and less flexibility. The Japanese auto makers, in particular, relied on a web of closely related but independent suppliers for many things the US makers had long ago brought in house.

Will we see more integration between Google and what are now independent suppliers? Almost certainly. Will we see Google take over the entire marketplace? No; and even areas that they do take over, they may lose in the long run anyway. What's the "right" level of vertical integration? No one has a clue - but the market will eventually sort that out.

-- Jerry

  Kevin Bedell [01.02.08 04:33 AM]

This reminds me a great deal of Microsoft's growth strategy from the 1990's onward -- they created value in their own company by absorbing it from other, smaller companies.

They repeatedly took features that others were charging for and delivered them free as part of Windows, or one of their other offerings.

Killing Netscape's Browser business (for which they were eventually penalized by the US Gov't) by giving away IE was a prime example of this.

They also worked hard to tie these services into their windows platform -- to literally design the core platform itself as a base with which they could absorb other technologies. I wonder if this 'architecture' approach has a parallel with Google.

The net result was that MS increased its market value (ie stock price) by effectively absorbing the market values of these new features -- with the small companies that pioneered the features usually suffering as a result.

My guess is that we'll see this strategy with Google for a long time to come.

One potential difference between MS and Google though, is the difference in the regulatory environment now. My guess is that the US Gov't would be much more willing to allow Google to get away with market domination in return for access to the information Google has about US Citizens (and other people worldwide).

With MS, the Justice Department worked hard to enforce laws against unfair business practices. But things are different now in our 'post 9/11' world.

Of course, it's also likely to be in Google's Shareholder's best interests for them to play along with the gov't. And given the fiduciary responsibilities of Google's management and directors, at some level that implies they should be open to giving the gov't what they want if that maximizes shareholder value.

I wonder how long Google can 'not be evil' when it's in their shareholder's best interests to turn to the dark side?

  Tim O'Reilly [01.02.08 07:31 AM]

Good point, Jerry.

You're absolutely right that it's not a technological issue. But generally, new markets susceptible to this kind of eventual vertical integration are born out of new technology. The automobile is exactly a case in point. In the early years of the 20th century, the auto industry looked a lot like the internet does today, or the PC market did in the 80s. Then as the market shook out, someone got big enough to do the integration.

You're also absolutely right that it's counter-productive in the long run. Microsoft is closer to home than GM. As Tacitus said of the Roman emperor Augustus (I think), "he made a wasteland, and he called it peace." Microsoft sucked all the life out of the PC ecosystem, driving innovation elsewhere (to open source and the internet), where people (originally) had no hope of making money.

Ecology rules in business as well as nature. Destroy your ecosystem, you eventually suffer.

  Michael R. Bernstein [01.02.08 04:04 PM]

I think some of this can be summarized as "there is more than one possible configuration for mature markets in dynamic equilibrium", noting that some configurations are simply more common, some are more prone to stagnation, and there can be disagreement over what constitutes the market.

Some examples are 'One Vendor to Rule Them All', the duopoly, 'Big Three', 'Snow White and the X Dwarves', etc.

The least common (because it is least stable over time) situation is where there is no dominant vendor(s), but in many ways the lack of stability makes it the most economically efficient at delivering innovation (I'll re-reference my perennial broadband example). Unfortunately, although businesses in such a market can be very profitable, that is *not* an outcome on which VCs want to bet, for obvious reasons.

  Niraj J [01.02.08 07:25 PM]

Tim ,

Your quote 'If Google no longer sends traffic to an external site, but instead delivers the data (and any accompanying ads) directly, they are winning and their former suppliers of destination pages are losing."

Its the way you look at it. If you treat WIKIPEDIA as a content distributer then WIKIPEDIA is actually in direct competition with Google , because essentially they have about the same mission as google. "To be the information source for everyone". WIKI does not author information they are content aggregators similar to Google in nature.

If google got into the business of page authoring like New york times or Wall street journal and started something like the Google Journal and gave a preferential page rank to the Google Journal then I would agree with your argument.

  len [01.03.08 06:43 AM]

Large companies that provide services over products tend to act like black holes in star systems. They speed up the orbits of near transients and cause them to emit noticeable signal, but then consume them swallowing all signal. Google isn't the first. IBM is noticeable in this regard as well.

Sometimes one SHOULD look a gift horse in the mouth. What one wants to watch is the content of announcements made to promote a service that potentially absorbs a client's market.

  Anthony D. Smith [01.03.08 06:57 AM]

An incredibly obvious observation is that capitalism tends to prevail in a capitalistic society. The ultimate winner is the holder of all the resources. I think I'll do a Google search to see how this article ranks. :-) Thanks Tim for the post. In addition to having a better understanding of Google's level of bias towards its economic gain, I'd also be very interested to know more about how that bias may extend into the political arena.

  Eric Blossom [01.03.08 09:40 AM]

I think that's ineluctable with an a rather than an i. Also the free download of Release 2.0 seems to be empty.

This may also be related. I find myself skipping to the second page or so of the search results list that I get from Google. I do this to avoid the heavy commercial pages that seem less pertinent. Places like Digg also seem to collect pointers to blogs etc. rather than to original sources. I'm clicking more as my eyeballs spin than I used to. Perhaps the golden age of search engines is already over from the user's perspective. They're still useful. Just a bit more painful.

  Web Girl [01.04.08 08:03 AM]

In some ways I think the introduction of google knol is a good thing. For one, I think that wikipedia is too easily modified by corporations, and special interest groups. With knol it may bring in a competitive atmosphere in which wikipedia will have to offer something it does not currently. In the long term I'm sure google wants to capitalize off the investment which could be a bad thing, or a good thing.

I will say that I use wikipedia quite often, and it does have great information, but I also find a lot of it to be wrong as well. Hopefully google will administrate knol a bit better then wikipedia currently does.

Overall I believe knol will be good for its users.

  Antonio Piccolboni [08.26.08 11:31 AM]

Sorry to be so late to this discussion. If Google search pointing to Google's properties is a problem, how about pointing to AdSense-running sites? This is hard to prove, but if you follow the link, you will see that AdSense-running sites get better ranking on Google than live.com, not based on anecdotes but hard stats. So one or both of them is using AdSense information to rank results, or please suggest an alternative explanation. It really doesn't matter to Google if they own a site or if it runs AdSense, more so in the scenario where Google search and AdSense achieve monopoly status and Google can wrestle most profit out of other sites. I wonder if I am the first to notice this, couldn't find anything so far. It seems to me a much bigger deal than knol, but part of the same strategy, and it's not in the interest of the user and the web.

  Gab Goldenberg [09.21.08 12:28 AM]

Yahoo shopping used to be based on affiliate deals and is now CPC, which is actually more profitable for them. They don't make $25 a click. It's $25 a sale. If you have 1/200 clicks actually buying, then you only make $1.25/click. Plus you lose the opportunity to sell the traffic elsewhere.

Yahoo's problem is that they try to partake in every part of the web, and in so doing excel at none.

  Yuhong Bao [03.04.09 05:37 PM]

"I wonder how long Google can 'not be evil' when it's in their shareholder's best interests to turn to the dark side?"
Well, IMO this itself isn't evil or illegal. What is evil is when the company tries to defend it's monopoly using tricks like vendor lock-in, like how Windows 3.1's AARD code tried to force Windows 3.1 users to use MS-DOS instead of DR-DOS.

Post A Comment:

 (please be patient, comments may take awhile to post)






Type the characters you see in the picture above.

RECOMMENDED FOR YOU

RECENT COMMENTS