« Popfly in your Face...book - Microsoft and Facebook | Main | CBS acquires Last.fm, eBay buying StumbleUpon »

May 26, 2007



Don- I think the analysis you do is is a very good one, but I think you are overlooking the network effect of dominant share in this market. You're absolutely correct that each percent of market share is valuable.. but I believe you are overestimating the value of each percent of share when your share is small.

One of the reasons why Google is able to generate the prices they do for their ad bids is because of the search traffic they bring to the advertisers. So in effect, it's not a linear curve from single digit market share up to the shares that Google owns. I'm sure there is an inflection point..

The same keyword on Google will absolutely be priced higher than the same keyword on a small share service. And of course there will always be a market for the lower priced good... but at the expense of lower margins for the company.

I argue that unless you can figure out a real game changer, fighting out each single digit of share against Google's established position is not the right strategy as you will always be stuck in a bad business where your margins cannot compete.


Great analysis Don - as I note on my blog this morning (http://www.sawickipedia.com/blog/2007/05/26/people-search-worth-30-billion/) it's easy to see what alternative and vertical search plays like Spock are getting so much interest. A billion is a big number and definitely worthy of all the VC money going into the space.

Don Dodge

Shan, That is precisely why 1% of search market share for Google is worth $3B, while 1% for the smaller competitors is worth *only* $1 Billion.

I did the math several different ways; bottoms up to show how you get to $1 Billion, and tops down with Google market cap divided by market share to show how they get to $3 Billion for each point of market share.

So, your strategy is to fold up the tent and go home because you only have 5% market share? Typical management thinking. That is exactly why I wrote this post...to open people's eyes to the huge revenues and profits available even for smaller competitors in search.

I agree that just slogging it out doing the normal things won't change much. But, any of the smaller players could make a few key strategic acquisitions, roll up some smaller local search companies, spin out their stock as a separate tracking stock, and then marvel at their multi-billion dollar jewel.

The private equity guys know how to create value. There are lessons to be learned there.

Emre Sokullu

Great post Dodge, this implies to even more (2x) valuation in 2,5 years with Forrester's numbers: http://emresokullu.com/?p=243

Jakob Nielsen

This is a good analysis. Two more points:

a) The value per page view on a content site tends to be about 0.1 cents. Thus, pointing people to content is more than 100 times as profitable as actually writing that content. (This is why I called search engines leeches on the Web, at http://www.useit.com/alertbox/search_engines.html )

b) Even though 1% of a huge number is a big number, it's a fallacy to believe that new search engines have any significant probability of getting a 1% market share. That's very difficult to achieve. Could be done, because GYMA are not anywhere near perfect. But it's not likely.

As to Shan's comment that bigger search engines should get a bigger CPC, that's not logical behavior from the advertiser's perspective. A customer is a customer, no matter where they come from, and you should pay the same for each click, whether it's from a search engine that sends you a lot or one that sends you fewer.

The one modification is that the conversion rate matters as well, and users from non-Google searches reportedly convert better than Google users, so most advertisers should probaby pay *more* per click from the smaller search engines.

Joe Duck

Fun and provocative math as usual Don. From a market perspective I remain surprised how the potential "instability" of the market is not weighted very highly. For example if PowerSet really does what they say they will be doing market share will shift within a few years (months?) to the new, superior search paradigm as it quickly shifted to Google back when they were the clearly superior choice for search.

Related: Why is MS failing so spectacularly in search when there is so much at stake??

Steve Morsa

Nice analysis as usual, Don.

Relatedly; any newbies hoping to take even a worthwhile 1% search market share from the others have; not just a tough battle to do so; but, actually, an almost super-human feat ahead of themselves.

Due to two critical factors; 1--the media's continuing love affair with Google (though we may be seeing some cracks in their "love fest" armor as G continues to make more and more controversial business decisions); 2--the (in my view unfortunate) fact that--for many in the US and much of the net-connected world--"Google" has become synonymous with "search."

As you and others have posited, Don, unless someone comes up with clearly and unquestionably superior search (AND is able to hold such an advantage for at least 3-5 years AND is also able to quickly and efficiently monetize it as well); tough with so many--including the incumbents--working so hard for better search...AND the media provides them with the same billions of dollars in free, usually gushingly positive press they gave Google (especially in their early years), there's little chance in you-know-where that we'll see anyone take even 1% out of the current search market for themselves; let alone 10-15%...

Even Superman would have a tough time racking up so many "ANDS"

Venlatakrishna Nalamothu

It is almost frieghtening to think about Google's monopoly over internet. Are the guys at Yahoo and Microsoft sleeping? In my opinion, by 2008, Google will have more than 70% market share and 90% of publishers are in the Google grip with recent acquisitions like Feedburner. If Google introduces Text link ads, 95% of ad market will be in the hands of Google by the end of 2008. Really frieghtening!


a very interesting analysis and to think just 1% of the market share already is worth over $1 billion alone.

Kaila Colbin

Excellent perspective, Don. I've been commenting about Jonathan Thaw's Bloomberg analysis indicating that a 10% rise in clicks for Yahoo would result in a 5% increase in revenues. We used his math to extrapolate that a 14% increase in relevance (the amount that MyWebDNA has been proven to generate) would result in more than $800 million for Google.

Your math comes in with almost exactly the same result: If each 1% of search market share is worth over $100M in revenues, then a 14% increase for Google, which has 50% market share, would generate roughly a 7% increase in market share. 7% x over $100M in revenues approaches the $800 million number we've been talking about.

These are, obviously, extraordinarily rough calculations! But they certainly serve to highlight the exponential power of small increases in relevance.

Best regards,
Kaila Colbin
VortexDNA Blogger


I'd like to take the contrarian view here and challenge your assumptions. In my opinion, the internet is to a large extent a "winner take all" game. To give you two illustrative examples, consider a publisher who has a choice between AdSense and another XYZ ad platform. Suppose it is known to all publishers that adopting AdSense would on average give them 50% more revenues than signing on to XYZ. What would the market share of the two be over time? XYZ won't have 50% less market share because they are paying 50% less. Rather, with time it will be tend to ZERO since all publishers would eventually sign up only with AdSense. Similarly, consider an auction network like e-Bay. Suppose you know that e-Bay has a 50% larger n/w than (say) ABC. Given the fact that buyers know that their chances of getting a better deal is that much better, what would the relative market shares tend to over time? Similar analysis as the previous one. And there is a term for this, phenomenon which is - NETWORK EFFECTS. In other words, big will become bigger and small will be wiped out. This is evidenced from the ever increasing share of Google of search revenues.
If you indeed want to prove your hypothesis on what each % point of the search market is worth, you should be using what that % is worth for each of the players. You will realize that a % point for Ask is not worth the same as a % point for Google. You cannot extrapolate the worth of each % share of the market for Google to other smaller players and conclude that it is okay for them to even take a smaller slice of the market. It may be true in certain specific cases (like say vertical search where click through rates for targeted audiences could be higher) but you certainly can't draw such a broad generalization as you have.


Arun, In the second paragraph I made the point "Google has 50 points of market share and a stock market cap of $150B, or $3 Billion for each 1% of search market share. Other competitors don't win the same revenues and market multiples, but even at the low end, 1% of market share is worth over $1 Billion." So, I think we agree on that. Perhaps I should go back and put that sentence in BOLD so that it gets noticed.

Let me expand on that thought. The stock market values each 1% of market share for Google at $3 Billion. Yahoo only gets a valuation of about $1.5B for each of their 28 points of market share. The second place player gets half the valuation of the first place player.

MSN Live Search, AOL, and Ask, are all part of much larger companies. It is not possible to isolate the search revenues, and stock market multiples, for each of them. My guess is that the third place player gets about $1 Billion for each point of market share, and the 4th and 5th place players get about $500 Million for each point. Just a guess.

That leads to my second point, made in the comments section. To realize the full economic potential of these search businesses it might be more advantageous to spin them out as separate tracking stocks, allow them to make a few acquisitions, and let them be totally focused on one thing....search.

My overall point, I think still stands. Even the small search engine players can generate huge revenues and are extremely valuable. Search is not a winner take all game.


Don, I think there are two separate issues here.
a. What is the value of having a % share of the search market? I think you have done a good job here of demonstrating that it is huge. The value per % point could be different for smaller (non-Google) players versus the larger (Google) players. Nonetheless, given the gargantuan size of the overall search market itself, even taking a small slice of that is of tremendous value.

b. But, the bigger question is whether the smaller players will even be able to hold on to their small sizes and the answer in my opinion as I reasoned above is - No. Google's march towards total market domination almost seems inexorable. And I have history on my side here- According to Hitwise, Google's share of search has improved from 58.33% in Mar 2006 to 64.13% in Mar 2007. This has been at the expense of ALL its major competitors. Yahoo - has dropped to 21.26% from 22.30%; MSN has dropped from 13.09% to 9.15% and ASK has dropped from 3.99% to 3.48% in the same period.

The reason in my opinion is not just technological but a combination of technological and economic. Google has entered a virtuous loop of - Giant user base -> Attracts advertisers -> Attracts publishers -> G Earns more $$ -> G Ploughs back more into business -> G widens technology divide -> Attracts more users. Extremely hard to break that kind of a loop when you are a minority player.

In other words, using the same investor metaphor that you've used, its a double whammy for her. Not only are the multiples for each % share of market less, the share of the market itself is under immense threat. Doesn't seem like a compelling story for someone wanting to invest in a competing search company to Google. Not too different a reasoning from whats driven Google's incredible run on Wall St ever since it got listed.

Cliff Johnson

You're mistakenly comparing Google's entire market cap to your estimated value of its US search share of 50%. Google's worldwide search share is probably about 75%. Yahoo has about 14%, Microsoft has about 6%, and everyone else adds up to about 5%.

If Google has a $150B market cap and 75% of the worldwide search market, the market is saying that each point of worldwide share is worth $2B for Google.

At the same valuation, Yahoo's search business would be worth $28B, and Microsoft's would be worth $12B.

Since Google is gaining share while Yahoo and Microsoft lose it, it makes sense that the per point value of search market share would be higher for Google than the others.

James Cooney

Arun: The situation you are describing seems to assume two things: 1. the smaller player does not differentiate themself from Google in any meaningful way, and 2. Google does not make ANY mistakes which could alienate customers. In the presence of those two events, I agree with you 100%. But I don't think they're likely, especially number 2. I still agree with you in principle that it doesn't seem like a highly compelling story for someone wanting to invest in a competing search company to Google, although for the right smaller investor, it might be somewhat interesting. Might, somewhat, smaller, these aren't words which get tons of attention, but meaningful revenue streams have been built on them. That said, it's not a business I would want to get into myself, and that perhaps speaks more than anything. Cheers.

Kaila Colbin

Hi again Don, and community,

As a group of people with an obviously informed perspective on the valuation of search market share, I implore your kind assistance.

There’s been a bit of semantic confusion about the validated VortexDNA results.

In an earlier comment on this post, I stated that a 14% increase in relevance would be worth $800 million to Google, according to our original calculations and corroborated by Don's Google valuation.

My basic premise, however, was wrong.

That 14% is an increase in relevance, not clickthroughs. What we have validated is this:

A Google user is 14% more likely to click on a search link with a high VortexDNA score than on a link with a low one.

We believe that the 14% would subsequently generate a 3% increase in clickthroughs for Google. Applying our original revenue correlation to a 3% increase in clickthroughs would result in $300 million in revenues.

Rather than commit myself further, though, I'll put it out there for discussion: any thoughts on what a 14% increase in relevance might be worth?

Thanks and best regards,
Kaila Colbin
VortexDNA Blogger

Arjun Thomas

Brilliant article.... makes you stop and think every time you run a search in google... 12 cents a search!! that really shocked me. Maybe i should start my own search engine :)

PJ at Ferodynamics

1. There's no "winner take all" Internet. You can't buy all the domains (I'll register more) and you can't control all the advertising. Google needs a diverse, interesting Internet as much as we do. Or nobody will use it. No matter how many Adsense units they come up with, we will always want more stuff to click on.

And you can't say Google controls x% of search--there's no way to measure it. We're always searching, all the time, on every website, even when you're not clicking one of the billion "submit" buttons out there to find something. Your eyes and mind search this page as we speak, and you're not using Google! The Google/Yahoo/MSN search box is just one piece of the puzzle.

2. Don't worry about everyone using Adsense. Once the match is made there's no way to stop publishers from talking to advertisers, and vice versa, even though the TOS forbids this. The technology is for us, not the other way around.

Also, publisher loyalty to an advertiser and/or network is not just about who pays the most. Think of the billions of unpaid links/endorsements. Most people are on here for other reasons. Surprise, there's more to life than money.


@ Jakob

Great point about conversion rate. Techies tend to use Google over other search engines (I'd imagine), and I'm sure they are more adept at ignoring ads or even using software that blocks ads.

Your point about advertisers' illogical behavior makes me nervous, though. Widespread illogical behavior does not exist according to classical econ, and, though it clearly does appear from time to time, classical economic models have been shown to work pretty well over hundreds of years, and it's usually wise to look for within-the-model causes before going elsewhere.

So, as others have mentioned, clearly the reason Google keywords are more expensive is because more advertisers bid on Google keywords than those of other search engines. Advertisers do this -- rationally -- because the fixed costs associated with creating an advertising account at a competing search engine may not be justified by the small number of surfers delivered by those search engines (especially for small advertisers, i.e. the long tail). Admittedly, those costs are probably perceived to be higher than they really are due to unfamiliarity with the technology. And Google's superior results are probably probably perceived to be even better than they really are because of Google's high press coverage. But even so, the behavior is firmly rooted in a rational cause.

David Scott Lewis

Powerset might be the next darling of MSM and the blogosphere. And if not Powerset, perhaps some other engine.

My point: Don't discount Powerset or any other truly innovative search engine.

Of course, at this point we (well, at least I) don't know how good it really is. But if Powerset is half as good as its PR, then it will be a formidable new player on the scene.

Assuming they don't do anything goofy with their UI (and I hope they have Jakob Nielsen as one of the key front-end consultants), their back-end AI technologies might change the market. Let's face it, even Google keeps touting AI.

We all know that NLP is the Holy Grail for this game and Powerset may have the chalice at hand. Yes, this crowd knows to use a "+" before an entry and how to do nested searches, but we're not the majority. The majority needs the best searches possible using NLP.

Alas, it's not just about AI or even neural nets. (The search engine firms could learn a thing or two from the robotics industry.) But at least they're a step in the right direction.


Hi Don,

With respect to the search market share post, I have a question. Google has some 7-8 categories(web,local, groups etc) where they display ads. If ads bring revenue, when comscore says 3.3B for google sites, is there a split-up of how much each category contributes to this total 3.3 B or is there an assumption that only the default web search contributes to the search clicks. if you can give some pointers, it will be helpful. I am trying to understand this diversity and its contributions better.


Thanks for giving hope to InfoCream, Mahalo and ChaCha.

trademark registration

Its amazing how profitable search engines have become recently. I wish I was able to invest in google back when it was still two guys sitting in a garage with an idea.


wow..imagine that's for only 1%...heck their making lots and lots and lots of money.. i should buy Google stock ...huhu..does anyone know how we can buy Google shares outside USA ?


On Friday a company with a market cap of roughly $78 million claimed a billion searches on its ValidClick AdExchange in December, (see link)

Is this an inaccurate claim or does the valuation process not apply to a smaller companies?

The comments to this entry are closed.

My Photo

Enter your email address:

Delivered by FeedBurner

Twitter Updates

    follow me on Twitter