A company making $100 million a year disintegrated when Google took away their business model. | Hacker News Reader

A company making $100 million a year disintegrated when Google took away their business model.

canadait.com

72 pointserik52 comments

Article in short:

  Geosign was doing "search arbitrage", which meant buying
  keywords from Google and filling their landing pages with
  ads from other providers (e.g. Yahoo).

  Google didn't like that. The end.

Man, can I hire you so I can surf about 90% less everyday?

Can anyone explain why the comment in this subthread by keating is marked as "[dead]"? What does that mean?

"dead" means "killed by a yc editor." somebody with some authority around here thought the comment was extremely inappropriate, trolling, spam, etc

Hmm, so why doesn't it just get deleted?

So people can't cry "censorship"

But it's still censorship?

generally speaking, they are. if you don't want to see them, set the "showdead" parameter in your profile to "no." i assume they're kept around in zombie state for forensic purposes.

That article took a long time to get to the point.

It was still interesting.

The purpose of feature writing is to tell a complete story, not provide data as fast as possible.

War and Peace in short:

  Napoleon invades. Gets kicked out.

THANK YOU! I could not even read the whole thing. I just got done watching 10,000 BC and guess I am not in the mood for another epic.

Seriously, can i get a feed of your Articles in Short.

Frikkin' A. that's the last time I read th article BEFORE I look at the comments. Could have saved myself 10 minutes.

I'm glad they got crushed, this title should read "Domain squatter nets $100 million a year scamming, until Google closed the loophole". Nationalistic pride aside, there is nothing about this company worth keeping around. It sucks for the employees who didn't know the big picture, but certainly not for the CEO or the VC, they knew they were playing with fire.

The company's business model had nothing to do with domain squatting. It is enough to have a single domain to do huge amounts arbitrage. Or should I say "it was enough". :)

By the way, there still is a number of smaller companies doing Google->Yahoo arbitrage with Yahoo's quiet approval.

It would be interesting to see what happened to that if Google bought Yahoo.

They could start arbitraging into second tier networks, like FindWhat or LookSmart. But it's unlikely that this will actually work in practice. The bids in those networks are too low.

Another way of doing it (and that's how some arbitragers are structuring it even now) is to have two or more separate companies, with different billing addresses and access respective Google accounts using different IPs.

granted, but arbitrage is a wonderful feature of capitalism. It shows you exactly where big inefficiencies are.

Value was : "search arbitrage" In essence, taking a profit out of google's adwords/adsense program while google kept fine tuning its pricing algorithm.

The term may sound fancy, but this business model has been questionable from the get go. I've read a few studies, and this type of practice artificially inflated prices on certain keywords, the process would start by these companies buying a few ads on the main google page(via adwords), if a user clicked on the them, they were sent to a landing page with more ads of the type(via adsense), in the hope's they would make up the difference clicking on more ads....that's were the 'arbitrage' came in, Google didn't have the pricing model right: to the tune of $100 million a year, its certainly isn't illegal, these companies just figured out what keywords were being paid more on the content side and cost them less to advertise themselves.

As an added note, this type of practice or 'search arbitrage' tends to beget more shady practices, since the revenue margin is so tight ( and its been getting tighter ) practices domain tasting and domain kiting, I wrote a more thorough piece on this whole subject: http://www.webforefront.com/archives/2008/01/advertising_con... ( Toward the end is a news link on how Google is cracking down on this type of 'search arbitrage' which more than often equals: domain tasting )

"And much of it revolves not around its plans for a content powerhouse, but the story of how the Guelph startup found a loophole in Google's vaunted advertising model, enabling it to make boatloads of cash - until Google decided enough was enough."

[...]

"What's more, he envisioned a network of thousands of websites all automated by software linking keywords to pages filled with ads, returning millions in cash in the process.

By 2005 that was exactly what was happening. Nye crafted a maze of Internet sites that included tens of thousands of Web pages and bought up even more keywords from Google. By connecting the keywords and the websites, Geosign was indeed generating more than $100 million in annual revenue and was extremely profitable."

Well, no sympathy there.

I liked this quote though:

"There's a tendency in the press to make everyone either a dog or a god. The truth is most people are somewhere in between. Tim is no different. He has shortcomings, but he's also a highly creative guy."

Yeah, exactly. The tone of the article is a little off. Fundamentally, these guys were advertisement parasites. They added almost no value to the internet, offered no meaningful products, and the world is frankly better off without them.

And yet I get the idea that we're supposed to be sorry for them because they're just a poor canadian company unfairly crushed by the american giant.

I especially like the use of the euphemism "search arbitrage" to describe their business model. :)

So you can say hand on heart, if you worked out a way to 'game' a couple of systems, in order to make money with little or no effort, you wouldn't do it? Just a little bit?

You could say the same about "direct to advertiser PPC" - people who use adwords to bid on keywords, then drive the traffic straight to other affiliate programs, and skim off the profit. Yes, on the one hand they are parasites, but on the other, they are doing keyword research, generating more sales for the merchants, etc. They are value adding.

Granted though, search arbitrage is not in the same boat, and doesn't really benefit anyone.

That's not what I said, though. The fact that I might be tempted to break a rule doesn't mean I'm morally obliged to shed tears for those who break it and get burned. The folks who were most hurt here were the investors, who clearly bought into the scam not realizing the risks involved. So if any tears are shed it should be for them.

But what really irked me about the article was the pseudo-nationalist tone. "Who cares if they were parasites? Those were canadian parasites that Google killed!"

They didn't break any rules though. They just played the system.

I agree it was written a bit like you say though.

The article clearly states that the investors did understand the risk. The folks who were most hurt were the newly-hired employees who had just quit their previous job to join Geosign.

As they should. The company did not add any value.

So the big lesson here is that you can make $100 million a year (for at least a few years) without adding any value.

So where do I sign up?

Seriously, I was concerned about it until I read the word "loophole" used. The only impressive thing the company did was bend/break the rules for so long and on such a grand scale without getting caught.

So when will they turn off the funds to the domain name squatters?

Not any time soon. Domain traffic converts well for many advertisers.

Just like someone can make a good amount of money playing Russian Roulette for money. But most who try that... and the risk increases with usage.

So wait - a scuzzy company got found out and shut down? Hard to be upset about that.

Good tale, however poorly written. Over half of the article told the story of the company. The story should have mentioned more about why it failed and it's rescue measures (the last third) instead of trying to draw the reader in emotionally to a business article.

Exactly... stop telling me little stories. I don't read an article like that for its literary value. Get to the point!

It's been a while since I've read an entire article all the way through. Good read.

Interested me because this happened about an hour from where I live.

Next up for the ax: gray hat SEO, where you suck in an RSS feed of stuff you didn't write and puke it into a Wordpress blog full of ads on a regular basis.

These types of sites kill Google's quality and I don't expect them to live long.

They're called "Splogs", and have been dealt with by Google already.

http://en.wikipedia.org/wiki/Splogs

welcome to 2006

The other more lucrative part of this is:

1. bid on high traffic, low cost keywords. 2. Use landing pages with content geared towards attracting High paying adverts 3. profit!

For example, you bid on "free stuff", pay 5c a click. Then send people to a page about class action lawsuits or something, where adsense is showing them some adverts for lawyers, with a $5 per click...

Similarly though, you'll get penalized for this now, as having a non-relevant landing page, and your bids will go up.

I'm glad they got shut down. Sitting in the middle doing nothing in a loophole isn't a valid business model for long.

Reminds me of SEOs who bitch whenever Google changes the ranking algorithm. Try a non-parasitic business model for a change.

Google-based revenues can stop with the tweak of a Google TOS.

This company's "business model" was even worse than Netscape's once Microsoft Internet Explorer forced Netscape to make their browser free to general public starting in January '98

Netscape revenues (browser only)

1995: $44.3M

1996: $181.2M

1997: $105.5M

1998: N/A

[http://www.usdoj.gov/atr/cases/f1900/1999/4.htm]

Not quite analogous, as Jim Clark always planned to make the bulk of Netscape's revenue off their server products. The exact quote was something like, I'm selling printing presses but first I need to teach people to read.

And then comes along Apache webserver!

Heh :-)

Back then Netscape Enterprise Server did a bunch of stuff that Apache didn't (threads, API, certs, etc). Most people were using forking and CGI scripts, but we were saving a ton of money on colo that easily paid for the license by using NSAPI instead of Perl. Even now, over a decade later, I maintain that NES 2.01 was possibly the best web server ever written. If I could use it today, I would.

What really killed Netscape was that the only thing less reliable than NES 3 was the watchdog process they included with it to restart it when it crashed! That was around the time that Apache was catching up, and the rest is history.

Search arbitrage based on geographic areas is pretty neat actually ...

Let's say you buy the keyword "mortgage" which is low value/high volume because it's broad and doesn't imply intent. You then geo-target your traffic to just people in California, and send that traffic to a page with "california mortgage" ads. Which of course are much higher value as they are more targeted and represent a smaller niche. Profit.

It's really quite brilliant. All done using Google's own tools.

Lots of comments about being a parasite, adding no value, and that he deserved it... but I don't see it that way. Sure, it was "obvious" that it would eventually get shut down, but he built a $100 million/year business out of it... and knew that it was going to be short lived. I respect his plans to flip it into a legitimate business. If Google had not shut it down for about 6 months more, Geosign might have become a major player.

I cant understand wtf the point of the article is. They make it seem as if the company was in some way picked on by the mighty google. Point being, they were committing FRAUD, don't let words like 'Arbitrage' confuse you. Its plain and simple fraud, you try to get around the system with dishonesty, you get shutdown. The real scary part of this whole article though was that VC FIRMS invested in it!!! WTF?

Where is the fraud? They bought legitimate keywords and sent the traffic to sites with more ads.

It's pretty much the business models of tens of thousands of web properties, just boiled down (no content) and refined (automated on a mass-scale) into something that makes them millions.

Hey man, if people are doing it, it does make it right. My definition of fraud is gaming the system as they did, and using it in a way that was not meant to be used to get $$. It happens in all industries and it can drive a company into the ground(close up shop)

The most stupid thing for a company to do is to get in the way of a giant, or for a consultancy to have a single customer.

Wasn't this predictable?

How many startups are out there that "do X like Y company, except our pricing model is better"? With such a subtle difference, really, shame on those who are taken by surprise when the giant pounces.

So... you haven't read the article.

I think he is just ranting in general.

This vaguely reminded me of the movie "Boiler Room".

If it sounds too good to be true...

And the book "Wolf of WallStreet", autobiography of the guy that founded that "Boiler Room"

Reply on news.ycombinator.com