Hacker News Reader

Comment by gilgoomesh | Hacker News Reader

If we take this article at face-value (that Google is as big as it is because of its monopoly in search) a related question is immediately raised: why is Google's search engine a monopoly?

If the monopoly is due to superior technology, why are Google able to write a better search engine and maintain this search engine lead?

If the monopoly is due to other effects (buying favored search-engine status in Safari, Firefox and pushing Chrome/Android) why don't the browsers have all the power?

I guess I agree with the article that Google's power is search-engine derived but there's more to the source of that monopoly than I think the article discusses.

View on HN

Good questions, and there are some reasonable guesses at the answers, I paraphrased your first one as:

   > How did Google get a search engine monopoly?

Technology, Timing, and Momentum.

Google exploited the social signal of linking to organize the list of web pages by those considered most valuable by the people linking to them. That relatively simple fact made their search results much better than the others. And of course they patented it (PageRank).

They arrived at a time when advertisers were just learning that they would have to advertise on "the net", and they offered advertisers something that other channels could not provide, feedback on views.

As the 'biggest' web site (by pageviews) they get the majority of the advertising dollars from advertisers, and they combined that with only showing ads about '<X>' when people were searching for '<X>' (fill in your product)

The environmental piece of the puzzle was that the number of advertising dollars is finite. A lot of people don't understand that. If you create a web site for someone to advertise on, new "inventory" as they call it, to get someone to pay to advertise on it you have to pull those ads from someone else's web site.[1] And because of that Google's success sucked a lot of oxygen out of the online advertising ecosystem. My first start up after Sun was a web magazine and the 'best' banner ads came from Yahoo who could pay you top dollar to show them and Alta Vista was the best search engine. When Google launched search advertising the value of a Yahoo! banner revenue went down by 95%. Like a Redwood dropping sap on the ground to kill off seedlings underneath it, Google's success was making it hard to succeed without it's ad juice.

Now everyone can do PageRank type search, and in blind tests Bind and Google results are indistinguishable. But momentum keeps Google on top. Interestingly, Facebook could take 10 or 20 billon out of Google's market cap by doing search in addition to social, but for what ever reason they have, they don't.

Your second question is thus:

   > Why don't the browsers have all the power?

And the answer is they do, and of late "Apps". Once Bing search results achieved parity with Google, people stopped switching "back" to Google from their non Google searches. Microsoft market share has grown on the back of IE doing Bing search by default when you type in the address bar. And now Mozilla has chosen to show Bing results (via Yahoo) as the default. It will be interesting to see how many people switch it back to Google. If you look at Google's quarterly results you will see they break out a number that is called "paid distribution." That is code for money Google pays for others to send search to their engine so they can advertise on it. Its gone from a miniscule amount in 2005 to nearly a billion dollars a quarter this year. Paying others like Apple to send searches on your iDevice to Google. I generally interpret that as a sign the monopoly is eroding.

The headline though, that Eric doesn't understand how Google works is incorrect. He knows exactly how Google works. But the book he wrote isn't about that, rather it is a "narrative" which describes things in a way that imputes knowledge and wisdom and greatness. The point is to make him and Jonathan look good in the eyes of the public and to tell a good story. I think it succeeds on both counts. How many people would buy a book about how luck and timing made people billionaires and in nearly 20 years of trying they still haven't come close to duplicating it? Imagine the autobiography of someone who won the lottery, pretty ordinary life, some personal triumphs, sudden massive wealth, a lot of sudden "friends" by people who would seek to influence that sudden wealth. Guess what, if you have a private jet and offer to fly friends to Cannes to catch the film festival, you will get your choice of people who will fly there with you :-)

[1] Yes, there is a infinite supply of 'remainder' ads, where people will pay 1 - 3 cents for an ad like stock brokers who put in open limit buy orders for .01 to catch to flash crash bounces. Those ads won't even cover the electricity cost of the page to host them.

With regards to Bing achieving parity with Google, Is it also true for deep searches ? because usually in deep search, searches you expand quite some time on getting to the answer, while trying quite a few queries, at some point Google often has a knack to understand exactly what you want(maybe with the help of your personalized data) and gives you phenomenal results. I haven't seen that kind of results from Bing, in my very limited experience.

So how does deep search with Bing compares to deep Google searches ?

And BTW, what are your search preferences, as someone who's probably an expert search ,and with so much domain knowledge(from blekko) ?

Google has invested a lot on the long tail, and Bing hasn't as far as I can tell. That is evidenced by achieving better results with longer queries. We've done some research (to be published soon!) at Blekko that has some similar benefits on the long tail although the index we keep is much much smaller than the big guys.

Thank you.

Google's monopoly is based on their early superior technology. That's it, pure and simple.

I'm old enough to remember the web before google, and it was terrible. Back then we switched between hotbot, altavista, yahoo and a few others - and they all sucked. Some of them ranked searches alphabetically - try searching through 100.000 results that are ordered alphabetically.

I remember the first time I ever used google. From the very first day I never used anything else, and recommended it to all my friends. If you haven't experienced the web before google I think it's hard to imagine what it was like.

Pagerank made the difference.

I remember the pre-Google search days too.

> Pagerank made the difference.

To search.

But they don't make money directly from search.

They were a superior technology that was doomed to go broke until AdSense and AdWords came along.

That's of course correct.

For the first few years they ran without an income model and hoped something would come along. That something was adwords, and it made all the difference. But an argument can be made that it was luck, google didn't start out with this income model, and kind of stumbled into it. A lot of people were initially opposed to the idea, but they tried it anyway, not thinking it would be a runaway succcess.

That's the thing, though.

Hoping something will come along is a terrible business model. Google actively sucks at every business except floating down the Mississippi of cash. Amazing technology and technologists, terrible business.

They could double their workforce, or hire only people with comparative literature degrees, or force everyone in the company to speak Esperanto on alternating days with Klingon, or fire nearly everyone. It would be dramatic, but in terms of the bottom line it would be pretty close to irrelevant.

Google is not a technology company. They are a cash pipeline operator.

Google is very good at building tech to support search. That's no small feat. But they're terrible at sales, customer support, etc.

People buy Google ads because everyone uses Google for search. So Google doesn't need to sell anything. People come to them.

It's an amazing position to be in, but it's a very unusual one.

> Google actively sucks at every business except floating down the Mississippi of cash.

> It would be dramatic, but in terms of the bottom line it would be pretty close to irrelevant.

I'm assuming you're talking about all of the people in the driverless cars group, android, etc.

Two things at work here:

1. As a business their size you either try to take every single dollar that floats down the mississippi or you make the river bigger. I would argue that pretty much every strategy decision that Google makes is to do those two things (example - driverless cars means more people using the internet - more people using the internet means more search - more search means more adwords. aka the mississippi gets bigger)

2. Ubiquity. Starbucks sucks, but it's on every corner. Google's strategy is no different.

I call Google's strategy the Spaghetti Cannon: fire spaghetti at the world and see what sticks.

Microsoft have done slightly better at it. But in terms of ROI, they might've done better by just parking it in index funds.

That is not quite true - they sold usage of their search engine to companies like Yahoo for a nice amount.

Unfortunately since this time results have gotten worse ( anecdotal I know, but I know many who agree). I think this is down to a mixture of issues - the SEO industry has made search optimisation much harder, googles probably been a bit complacent with no competition too. Also introducing big changes like search personalisation and localisation which (IMO) has taken away more than it adds.

Well indeed, Google versus SEO is continuing fight.

But point is that Google has ante'd up to maintain search as something that works easily, that finds what you want, etc., even while against countervailing factors.

And bringing it back to the parent, the monopoly Google maintains is a dynamic process - they aren't always fighting against other search engines. Sometimes they are doing that and sometimes they are fighting against SEO and spam, sometimes they are competing with walled gardens and whatever forces would turn the Internet into an appliance, sometimes they are fighting to make search more compelling. But they don't fit the profile of a company just sitting on a monopoly even if they maintain is what mostly brings them revenue.

Unfortunately since this time results have gotten worse ( anecdotal I know, but I know many who agree)

Showing results for I know many who disagree. To search for what you typed in and wanted to search for, jump through {this hoop}.

It's worse than trying to write a paper on the Teh tribe in the Amazon in Microsoft Word with autocorrect. The relationship between query terms and results is complex and opaque.

That's to some extent by design. When organic search works perfectly, the user goes directly from the search result to their desired destination. The search engine makes no money from that. This is the fundamental conflict of the search engine business. When Brin and Page originally tried to sell their technology to Yahoo, Yahoo rejected it because it was "too good" - users would immediately leave Yahoo.

Now Google has that problem for itself. About a third of Google's revenue is from AdSense (non-search ads). It's not in Google's interest for their anti-spam efforts to be too tough on made-for-Adsense sites. They crack down only on the outrageous cases. This became very clear in the "Sportsdrugs.com" case, where Google was caught in an FBI drug sting and paid $500,000,000 to avoid prosecution.

If you want to understand how Google really works, read the black hat SEO forums. Every time Google tries something new in search ranking, it's reverse engineered in a few weeks. Trying to detect SEO with machine learning based on page contents worked for a while, but now the SEO guys use machine learning too. "Automated SEO" tries to automatically reverse-engineer Google's anti-spam algorithms, then adjusts pages and links until the SEO's model of Google says the page is under the threshold where Google will consider it spam.

It's not that hard to detect web spammers, if you're willing to do a quick background check on the company behind each web site. This can be automated; see our "sitetruth.com". Large numbers of spam domains trace back to the same source, so dealing with the problem on a per-business basis rather than a per-domain basis works.

Google has a monopoly on mind share, but only about 2/3 of US search. Bing (including Yahoo, which resells Bing) has about half the search volume Google does. That's historically about where Chrysler has been relative to General Motors. Yet Chrysler is taken seriously as an competitor to GM, while Bing is not taken seriously as a competitor to Google. It's not clear why. There's no network effect in search; it matters if your friends are on Facebook, but not if they're searching with Bing.

Bing has a very low public profile. It's hard to even find out who's in charge there. (Searching with Bing won't help.) Previously, it was Satya Nadella, who moved up to being CEO of Microsoft. There is no Bing CEO any more. Bing is under Microsoft's Online Services Division, headed by Qi Lu. Different parts of Bing are under four different parts of Online Services. (The Microsoft memo: http://www.geekwire.com/2014/internal-memo-microsoft-sets-le.... If that hadn't leaked, nobody would know who was in charge over there.)

Pagerank was indeed a giant leap at the time, but some other things are frequently forgotten with that one-trick pony narrative.

* Ownership structure - Google's founders had the front to IPO with a structure that gave them complete control of the company (10:1 voting rights over ordinary shareholders). That alone is why they can make huge long-term bets, from GMail to Android to robotics, instead of dumping inventory on the last day of each quarter to appease Wall St.

* Ads - Colossal ad revenue wasn't a guaranteed point on the trajectory from Pagerank and should therefore be considered as an independent innovation. They initially monetised with the search appliance for intranets and there was a time when they were considering a charging model for companies to use Google search.

* Superior ops - Google had to scale like crazy. The early crawlers were supposedly using 25% of Stanford's bandwidth after much pleading with admins, and the scaling never ended. Other startups might have partnered with some giant hardware company or bought into various commercial Unix offerings, but Google's early team figured out how to operate and distribute giant data centres on budget.

> I'm old enough to remember the web before google, and it was terrible.

Yep. I worked for a newspaper when AltaVista was the best thing going and I'd help the librarians and research team craft queries to get information based on understanding the finer points of AltaVista. When the Google beta was available I showed it to them and was never called on again. It was like magic compared to what went before.

That's not a sufficient explanation. It has been more than 15 years since they launched.

I would argue that their monopoly isn't due to that initial insight but a) their relentless improvement of search over the years, b) their technology infrastructure that made large-scale computing much more manageable, and c) their early lead in dramatically reducing computing cost, allowing them to spend dramatically more computing power per search than their competitors.

I'd add that another important choice was ignoring the modern business dogma of revenue maximization. They kicked the hell out of the early search advertising market by limiting things to well-chosen little text ads, something that was initially lower revenue but delivered much higher value to the users.

> Google's monopoly is based on their early superior technology. That's it, pure and simple.

I'm sure you know that monopolies are illegal in the USA?

Also, Google's search engine is not the sole reason for this monopoly. There's a political side to it which most Google fan boys prefer not to discuss.

> I'm sure you know that monopolies are illegal in the USA?

Not true: http://www.girardgibbs.com/antitrust-monopoly/

No, monopolies aren't necessarily illegal in the USA. But they can be depending on the behavior of a monopolist.

From http://www.girardgibbs.com/antitrust-monopoly/:

> Monopolies are illegal if they are established or maintained through improper conduct, such as exclusionary or predatory acts. This is known as anticompetitive monopolization.

Pagerank definitely made a difference - but I've always argued that the speed of Google results also was a major factor in winning over consumers (and something I rarely hear people mention). I remember the first time I saw results from Google, I simply couldn't believe they could have delivered them that quickly.

I think Google's success can be attributed to the "4x factor" - that is, that in order to change consumer behavior, you need a solution that is not the same or twice as good as current technology - you need a solution that is, at minimum, 4 times better. That's what Google Search delivered (and the iPhone, Uber, et al).

So for me it is easy to understand why people don't switch to Bing or whatever - their results are maybe, at best, 1.5x as good as Google - not good enough to alter behavior.

I would say that it was technology + design, rather than technology alone.

It's also very important to remember that Google also looked different than Yahoo!, AltaVista, etc. They were large portals with huge numbers of links. It was pretty confusing. Google was very simple. There was a search bar on a clean white screen.

This point is really hard to overstate. I'd go so far as to say the majority of the usefulness of the modern web can be attributed to Google making it usable. It was like adding power to a circuit.

> If the monopoly is due to superior technology, why are Google able to write a better search engine and maintain this search engine lead?

Answering because I feel like this is the bulk of the answer -- Adwords still constitutes a large chunk of Google revenue, and depends directly on how good the search engine is. As for the search engine itself, they take a ruthlessly scientific approach to improving it: strict measurement, hypothesizing about improvements, testing the hypotheses, and then some world-class implementation and quality monitoring. This is generally evident in the level of rigor seen in Google-published systems papers.

I actually think the algorithmic search engine era is coming to a close, and that Google could quite easily lose its market dominance. Google search is 10-year-old technology that doesn't give good answers to the most popular searches. I wrote a longish post about it a few days ago, if you are interested. http://newslines.org/blog/googles-black-hole/

> why is Google's search engine a monopoly?

It's not a monopoly.

However, it has a massive network effect that has made it so far impossible for anyone to make a successful frontal assault on the two core businesses of AdWords and AdSense.

I think some people have redefined "monopoly" to mean "successful business."

People are also giving Thiel credit for this insight. Michael Porter was probably the first to look over the fence of the Economics department, find out what the abusive behaviours of monopolists are, and to begin to teach them as strategy to be emulated.

> impossible for anyone to make a successful frontal assault on the two core businesses

Well, it's kinda definition of monopoly, which Google of course is.

Except that there are hundreds, maybe thousands, of niche advertising platforms, markets, middlemen, services, systems, technologies, clearinghouses, inventories, sellers, buyers, brokers, verifiers, subverters ...

The advertising ecosystem is actually very complex and diverse. Nobody is quite sure how big it actually is. But Google definitely has the two biggest chunks of it and nobody else is able to, for example, place keyword-based ads into Google's search results without Google taking their cut.

It's a monopoly because of multiple factors (some of which you've expounded on already.

A main reason Google's search engine is a monopoly is due to switching costs and the fact that all of our habits are built around using it.

Another way of putting it: once the word google became a verb they had solidified the monopoly.

Reply on news.ycombinator.com