I’ve learned something very important in a project I’ve been working on this week. You get what you pay for. Never again will I purchase a hosting package at 1and1.com. Between problems transferring my domain, to problems importing a database, even to problems setting up a database- How can it take three hours to create a freaking database? I’m talking “create database” here!
This is the last post I’m making on arbitrage. I think my focus on it has been unhealthy and I’d rather encourage myself to pursue more interesting ways to make money.
I admit I took these steps primarily from this crazy experiment.
You’ll want to find a spot within these steps where you can shine. You can go down the steps here and look for what resonates with you. Keywords; do you know where or how to get keyword lists, especially with search statistics? How about Domains– can you register domains cheap enough to make your riches by selling domains? Or do you have the knack (and funds) for registering a bunch of domains regarding an up and coming hot topic? For instance, LinkedIn, the social network, is getting an IPO next year; what if you bought LinkedIn domains and speculated (like people do with penny stocks).
Are you any good at writing? How about churning out some content? And do you have the savvy to run a hosting environment? Maybe you could hunt out a niche of people who need hosts? ( take a look at adsense hosting for instance).
Here’s the deal. I don’t think you’re going to make a killing at the Google Arbitrage game.
But everyone else thinks they will.
If you’re going to make money from Google Arbitrage, the first thing you need is a niche. Suppose you had a means for generating a list of keywords and how much search traffic they had. You could sell access to your list. ( To be honest, that site has a free offer as well.)
You probably also want some advice. Jonathan Leger is a programmer who has done some work to “help other people make money.” His three-way-link site is a cunning way to help people get better search engine love with their little niche sites. Which would lower the cost of advertising (in a roundabout way). A lower cost of advertising is the other thing people who want to make a killing at Arbitrage need.
Yesterday, my example paid me $35 in a month ( gross, before I subtracted the $300 in advertising ). An account on the three-way-links site pays Jonathan $47 dollars. His advertising costs ( I imagine) are low. However, imagine if he sells 100 accounts. That’s $4700 each month. And it’s recurring; he will get this money for quite some time, and he can stop paying the advertising and still earn the money.
No, I don’t know how many accounts he has, and no, I don’t know how much he’s spending on advertising.
I’m not picking on Jonathan. His sites and his techniques are extremely smart. Nor am I picking on the nichebot guys; they’re doing the same thing I’m telling you to do. In fact, I think they’re more honest in some ways than most arbitrageurs.
When I was a kid, there was a comic strip ( I think it was spy vs spy in Mad Magazine) where two people got into a pie throwing contest. Anne bought ten pies and threw them at Bob. Bob bought ten pies and threw them at Anne. Both people wound up covered in pie, and both people had fun throwing pie at the other person. But who “won”?
The pie merchant.
If you want to make a killing from Google Arbitrage, my advice to you is to be the pie merchant. Find something that people who are trying to make money on the internet need, whether it’s keywords, links, hosting, templates, project management, whatever. Find that thing and sell it to everyone.
Arbitrage is taking advantage of the difference between two financial markets.
For instance, suppose I go to Clickbank and find an ebook that sells for ten dollars. The seller gives the advertiser a 50% cut, so if I steer people to his site and they buy a book, I make five bucks. Of course, every person who goes to his site won’t buy the book, suppose the “conversion rate” is a high five percent (two percent is normal). So for every twenty people I send to his site, I make five bucks.
Now suppose I decide that people just aren’t coming to my site fast enough, and I’m only making five dollars a month. I want to get more people coming in, and so I go to google adwords and set up an account. My ad program will pay one dollar per click for each person to come to my site. I set up the plan with a daily budget of $10. So I get ten clicks a day from this program. The user “converts” on my site to the seller’s site; assume I do a fantastic job of selling the seller’s stuff, and five people per day go to his site. It takes me four days to make one sale ( I make $5 ), and in four days I’ve spent $40. I’m $35 in the hole!. By the end of the month I’ve spent $300 to earn $35 dollars.
To make money from arbitrage, you need two things; one, a product that converts very well, and two, ads with a very low cost. As you can imagine, these things are a premium; if a product converts very well, everyone’s trying to sell it and the cost of the ad skyrockets. If the cost of the ad is very low, the product won’t convert very well.
Google Arbitrage uses the Adsense advertisements as the “product that converts very well.” Basically, you build a five page site stocked with keywords and hook up adsense advertisements to it. Now, you have a site that should get good pagerank, and if you “SEOize” it right, you should get some nominal traffic. Second, you purchase adwords advertisements ( remember, you need a low cost advertising niche ), and send traffic your way. If it’s done “right,” you will earn more money from Adsense than you would spend on adwords.
I think Google Arbitrage is legal. I’m not a lawyer, so take that as the opinion of some computer geek. Basically, you’re gambling that your site is good enough, your adsense costs are low enough, and your conversion and payouts are high enough to make it worth your while.
I’m not sure it’s “honest.” Most of the sites people set up for Google Arbitrage are scraped from other sites, poorly done clumps of keywords, et cetera. Some poor schmuck worried about his aunt Marjorie searches google for “High Blood Pressure” and finds a site dedicated to high blood pressure– it’s got articles someone scraped from the Wikipedia and it’s heavily adsensed. While he’s looking at it, he clicks on one of the adsense ads and the ( and here’s a nice fancy word for you ) arbitrageur makes fifty cents. Did he get the information he needed? I hope so. But my main beef is that the ( here’s that word again. No I don’t know how to pronounce it ) arbitrageur isn’t adding anything to the mix; there’s no actual new information on the scrape site.
I’ve heard that Google is looking at how to stop arbitrage like this. I don’t think they will. I don’t have any hard numbers but I’m sure there’s more people either making small amounts or losing money betting on arbitrage than there are people earning the five or six figure payouts that certain people advertise. I think arbitrage is a cash cow for Google. And to emphasize the point; if you have a google Adwords bill for $25 dollars, you pay it that month, right away. If you have a google Adsense “earnings” of $25, you don’t get it right away — you get it after you total (currently) $100 ( they don’t pay under a hundred bucks if you want a check ). And they have a month between when you break the hundred dollar barrier and when they cut the check to verify their results. How many of us would like to borrow $25 a month for four months from a hundred thousand people, and then pay that money back at the end of four months? Even if you break even, Google wins.
I’m not complaining about Google. I run Google adsense everywhere. 
I just don’t think that Google arbitrage is a very good way to make money. It’s not smart and it’s not safe.
I have some ideas on better ways to make money from Arbitrage though, and I’ll talk about them tomorrow.
Here’s one method I’m using to track down phishing sites that have been uploaded to a hosting provider I maintain.
All of the directories I’ve had to delete or remove begin with a dot (to hide from a regular ls listing).
It’s a php commandline script.
Modify the variables at the top of the script.
< ?php
$findcommand = "/usr/bin/find /home/*/www -type d -name .\* ";
$dest = '/root/found_hacking/';
$adminemail = ** YOUR EMAIL ADDRESS **;
$tar = '/bin/tar';
$rm = '/bin/rm';
exec( $findcommand, $current, $e );
$disallowed=array();
$today=date("Ymd_hi");
$host=`/bin/hostname -s`;
foreach( $current as $foo ) {
switch( basename( $foo ) ) {
case '.svn':
case '.htpasswds':
// those ones are allowed
break;
default:
$fn=substr( basename( $foo ), 1 );
$cmd="$tar cfvz $dest".$today."_".$fn.".tgz $foo";
exec($cmd, $trash, $err );
if ( $err === 0 ) exec( "/bin/rm -rf $foo", $t, $e );
$disallowed[]=$foo;
}
}
if ( count( $disallowed ) ) {
mail( $adminemail, "disallowed dot-directories found in web server ( $host )", implode( "\n", $disallowed ) );
}
I’m working with multitail and guardian… here’s a sample config file.
#snort log
colorscheme:snort
cs_re_s:magenta,,bold|magenta:***insert a regex for your protected ip list ***
cs_re:green:.*Priority: 3.*
cs_re:yellow:.*Priority: 2.*
cs_re:red:.*Priority: 1.*
colorscheme:guardian
cs_re:blue: MS-SQL.*attempt
cs_re:magenta:Odd.. source.* done\.
cs_re:yellow:WEB-CGI .* access
cs_re:red:Running.*
cs_re:green:expiring.*
# Default colorschemes
scheme:postfix:/var/log/maillog
scheme:syslog:/var/log/messages
scheme:snort:/var/log/snort/alert
scheme:guardian:/var/log/snort/guardian.log
A compatriot asked me if I thought this was a good idea for one of my clients: Barracuda Networks Spam Firewall. My response was a reserved “yes.”
Simply put, my clients historically have been cheapskates who would rather spend $25 an hour on a developer than $2000 on a specialized server. “Just slap linux on that and fire it up!” I have heard time and time again. And so my response to him was: “Why not just a dedicated SpamAssassin box? He liked the web-based user configuration, and the snmp and the dedicated logging; in short, he liked all the bells and whistles that Barracuda has built into their system.
When I encouraged him to look into Using SQL with SA, he agreed, and then we started dickering about pricing and how to “sell” this package. I, for one, as a PHP developer primarily would code a web-front-end to the SA tables in PHP, and I can see it taking me longer than 40 hours. And I, for one, could configure snmp and some logging on this box, but it might be tricky, and then I’d have to go back and document it all for the person who came after me.
Honestly, with the number of thrilling projects I have going at any given moment, it would be an easy sell to me for a $2000 piece of equipment that I could set and forget; as long as it saved me at least 60 hours of work. However, the client in question also employs an army of programmers in India. I have no idea what he pays them, but I’m sure it’s less than $25/hour. I don’t know where the break-even point would be for that.
So; does anyone out there know of a web front end to this spamassassin-by-sql thing?