« Irish thoughts | Main | The War with Mexico »
Google bidding games -- markets in everything?
What would happen if MicroSoft or Yahoo or a MicroHoo went to the 5 top results for the top 25k searches and paid them to leave the Google Index?
A theoretical maximum of 125k sites, but with overlap, probably closer to 100k or less, times how much per site on average?
The math starts to get interesting. At $1,000 per site average times 100k sites, thats only $1 Billion Dollars. The distribution would obviously favor the larger sites, so of that billion dollars, would the top 1k sites take 500k each and the remaining 99k split the rest?
Given the stakes, why stop at $1 Billion Dollars ? Would the top 1k most visited sites take a cool $1mm each, plus a commitment from Microsoft or Yahoo to drive traffic through their search engines to more than make up for the lost Google Traffic. After all, once consumers realized that Google no longer had valid search results for the top 25k searchs, that traffic would most likely go to Microsoft and Yahoo...
What would it cost to get that number of sites to turn Google off and stay off, and would the traffic created as users switch from Google more than compensate for the cost?
Or would Google recognize the risk and jump in and offer more to websites to stay?
That's Mark Cuban, but I say no. Is there any precedent for successfully undermining a popular "monopoly" (in this case a free one) by paying the input suppliers to go take a hike? If bidding ever did get underway (unlikely), those suppliers are worth more to the relatively efficient company, which is probably Google and they are certainly less liquidity constrained than just about anybody else.
Yes, Google does reap profit from the free content supplied to the web, but the chance of content suppliers colluding to keep reap more of those gains are small. If anything the relevant externality is that Google could pay for gated content to be accessible to search engines but I wouldn't bet on that happening either.
Posted by Tyler Cowen on June 15, 2008 at 02:58 PM in Web/Tech | Permalink
Comments
First, this sounds illegal as all hell. To make your bribe money back you're going to have to raise your ad prices eventually. This sounds either like a cartel or predatory pricing.
Second, $1,000 dollars for the biggest sites in the world? A vast amount of web-based business comes in via Google. A site I was involved with got dropped from the Google index for a couple of days and it was a major crisis. How many sales leads for, say, Ford dealerships are going to go missing in the first month till people realise they can't find things from Google? Big companies spend millions on their web presence: they're not going to halve their traffic for a thousand dollars.
Third, from Yahoo finance, Microsofts market cap is 270.74B, Google's 179.51B, Yahoo 32.29B. It doesn't make sense for Yahoo to get into a bribe war with a company much richer than them. Microsoft maybe... but Google's probably more highly motivated than them.
Posted by: TheophileEscargot at Jun 15, 2008 4:05:58 PM
Interesting speculation from Cuban, but I agree it's not very likely, though there are some approximate parallels in the mobile phone industry. US$1 billion is the general price for a pharmaceutical company to bring a major compound to market, and the spending war there is coming to an end as their profit margins erode. It's not hard to imagine a series of events that would provoke a similar situation in search: top sites wanting more, shareholders and analysts losing faith in the operating model, (gasp!) government regulation of information assets. The interesting bit to me in this is the possibility of a company or companies not only competing, but actively seeking the destruction of another company. Any recollection of anything like this happening in the past?
If something like that were to happen, and if I were a small player in the industry, I would look for some sort of workaround to exploit, essentially replicating the early days of Google.
Posted by: Joseph Logan at Jun 15, 2008 4:07:45 PM
Can sites stop Google from indexing them? By convention, search engines ignore sites with a robots.txt file asking them to do so, but I don't think there's any legal obligation to abide by this convention.
Supposing that there were a way to stop search engines from indexing sites, and assuming that Google decided to engage in a bidding war with Microsoft, would this be a bad thing? The primary effect, it seems, would be to transfer revenues from search companies to content providers, shifting resources from search engine development to content creation.
On the other hand, having to use multiple search engines to find what you want would be a pain.
Posted by: Brandon Berg at Jun 15, 2008 4:08:28 PM
First, someone can't multiply. $1000 per site times 100K sites is actually $100M, not $1B.
Secondly, $1000 isn't even close to what it is worth a top site to remove itself from Google permanently. For many of those sites, any given day of Google traffic drives more revenue than that, and possibly any given hour.
Third, the collective action problem here is fierce. Who exactly is going to want to take $1000 to screw up their business, when the only way it gets un-screwed-up is if enough other businesses take the $1000 so that MicroHoo beats Google? Little upside, enormous downside, for a long-shot bet whose outcome you can't actually impact noticably. Sorry, find another sucker.
Posted by: Dave at Jun 15, 2008 4:51:48 PM
As previous commenters have already pointed out, the amounts of money proposed by Cuban are just preposterously small. If he added merely one extra zero to the amounts it would already be economically infeasible and yet still not even close to enough for sites to bother with. Does Mark Cuban even use the Internet anymore?
Posted by: at Jun 15, 2008 5:25:20 PM
Top 5 placement in Google on a common search term is worth much, much more than $1000. Given how much ad revenue my company receives from Google referrals alone (we're not a particularly popular site but make significantly more than that per day on ad revenue), I can only see this scheme being prohibitively expensive.
Posted by: Stephen at Jun 15, 2008 5:28:03 PM
The main thing that Cuban's missing is that consumers would backlash HARD against Microsoft/Yahoo (as well as any sites that took their offer) if they did something like this. Google is a household name with GREAT products, and (like Apple) they have hella dedicated users. Trying to undermine Google in such a shady way would induce wrathful backlash.
Posted by: Andrew G at Jun 15, 2008 6:12:45 PM
1. The top sites by this measure are going to include a lot of publicly traded companies: the major news outlets, movie studios, apple, microsoft, airlines, hotel chains, the NFL, NBA, etc., not to mention non profits like wikipedia and other Internet companies. A valuation in the billions seems low.
2. Microsoft and yahoo have already revealed their preferences against this, since they allow themselves to be indexed by google.
3. The value added for a search engine is in the long tail anyway. If the big sites above were sufficient AOL and prodigy would rule the net.
Posted by: dk at Jun 15, 2008 7:46:00 PM
Disclosure: I work on the Wikia Search project. That's where my take is coming from.
Putting aside the other problems cited by fellow commentators, I wonder how much Cuban looked into how Search works before blogging. There are two fundamental problems.
1) "Universal search" is all the rage. That means search results don't just return relevant websites, but also other information. When people search for "Britney Spears" on Google, how many of them are looking for her official site, fan sites, her myspace page, IMDb entry, Wikipedia entry, etc. They're looking for news. And that's, increasingly, what Google features.
2) For the important sites, $1,000 isn't even close to enough. Heck, $1mm isn't enough. I know other people said this above, but there's more to it. Users tend to use Google for site-specific searches. You've probably done it by appending "wikipedia" "youtube" "imdb" "webmd" or something else to your search phrase. If one of those brands (putting aside the fact that youtube is GOOG and Wikipedia a 501c3) were to lose this benefit, even temporarily, the loss would be potentially fatal. WebMD is a $2bn company (see this with roughly $40m in earnings. $1m isn't going to get them to sneeze.
3) Wikipedia is a non-profit dedicated to the spread of free information. (NB that Wikia and Wikipedia aren't related, and I have no knowledge of how Wikipedia works, politically, beyond that which anyone could read about.) Wikipedia entries also appear atop many "fat head" searches. Being a non-profit, Wikipedia lacks the incentive to have itself bought out of Google results. So I'm pretty sure that the top 5 sites per top 25k terms would only change marginally.
4) How would Yahoo or Microsoft identify the top 100k sites per Cuban's methodology? I assume they'd use the data they collect via their own engine. Well, they're competing with imperfect information, because it's likely if not certain that Google's top 100k would be different, at least in order, than Yahoo's or MS's. This is significant because GOOG, with it's incredible free cash flow and market cap, could rather easily target some of the more valuable but affordable sites, and kill off the other engine's goal. That is, they'd not have to play Cuban's game. They could thwart it by buying up a few key properties. Heck, they already own YouTube.
Posted by: Dan Lewis at Jun 15, 2008 9:52:05 PM
I would agree with this analysis of Cuban's question if it were strictly true that the top 25k websites were input suppliers only. What makes things more complicated is that they're also customers of Google's Adsense and Adwords services. Although I can't think of any bribes of input suppliers working on a large scale. Customer "bribes" have worked very well over the short-term in grabbing market share. But they have also worked very poorly over the long-term in promoting an industry structure that offers a sustainable package of quality goods or services. The airline industry is an example of what happens when everybody tries to compete on price for marketshare. It's a total mess for investors and (after the wild price war is over) customers alike.
Posted by: Michael F. Martin at Jun 15, 2008 10:01:04 PM
This site is in the top 100k in traffic. Would you remove yourself from Google for $1000.
http://www.alexa.com/data/details/traffic_details/marginalrevolution.com
Posted by: joan at Jun 15, 2008 11:02:41 PM
Another problem with removing the top 5 results for the top 25 searches is that most of the content on the internet is repeated. A lot.
Also, most queries give you a Wikipedia entry or two in the first five results. How are you going to buy off that behemoth (as I recall, Google provides money and machines to the Wikimedia projects. As do Sun, and many others. But Google can always afford more.)?
Posted by: grant at Jun 16, 2008 2:30:12 AM
Is there any precedent for a search engine to dominate the market for more than a decade?
I don't think search, per se is what Microsoft or Google are concerned about. Microsoft is probably concerned about Google invading their operating system and office tools space, and almost any amount of losses on search would be worth the gamble as long as it hurt Google as an on-going business.
Rather than bribes, what if MS just bought the sites outright? Then they'd get the sites, the eyeballs, and the web development teams to boot.
The question is, are we already at the point of diminishing returns for the quality of search. I say no, because I search for moderately obscure stuff and have a really hard time finding it. However, I don't know for the average person. Would a minor decrement in search quality by eliminating the oligarchs, which may not be a decrement in the long run, actually make people switch? The whole idea of a brand is that once established you don't have to be the best and you still retain customers. This, by definition, is difficult to predict.
Cuban needs to keep his eye on the ball, and bribe Steve Nash to come back to the Mavs.
Posted by: Andrew at Jun 16, 2008 4:18:41 AM
This isn't so much an issue for Yahoo, but Microsoft should worry more about people blocking their search because it uses excessive bandwidth.
Posted by: Jim Hu at Jun 16, 2008 11:55:06 AM
There is a lot of debate in adland about where a Google-killer may come from. One possibility is to create a search engine which pays the *searcher* a share of revenue. Another model is for a proportion of revenue generated by the searcher to be hypothecated to a charitable fund, and each searcher to vote annually on to cause to which that money should be donated, with votes proportionate to the revenue earned by each individual. For all of us to have a vote annually on where 10% of Google-sized revenues could be spent might be more interesting and rewarding than to merely pay us each something dull like $27.42.
Posted by: Rory Sutherland at Jun 16, 2008 5:53:14 PM
