I got some Christmas money from my parents this year, like I do most years. Having recently begun tinkering with applications on the LAMP stack(Linux, Apache, MySQL, PHP), I grabbed a web server and a URL or two to play with.
My wife’s blog has done exceptionally well by my standards- over 30,000 hits a month. I bought her a WordPress SEO(search engine optimization) book, and she saw a significant increase in traffic after making some modifications to her posting habits and management of her site.
This got me really interested in SEO. With recent changes in how Google suggests searches and shows results in real time, it’s clear that controlling what information shows up in a search result and what information does not is a powerful tool to have. Suggesting results does even more to limit the depth of the average search, because it eliminates some of the inconsistency in how people search for items. As someone immersed in the hacker ethos since I was a 10 year old hacking AOL on a 14.4 modem, my first thought was: “how can this be abused?”
It’s not a popular opinion, but after my experience with SEO, I consider the entire field to be “greyhat.” Take a look sometime at keywords used in the sites of competing companies- it’s all too common to find one company deliberately parasitizing the traffic of a competitor. So many sites are deliberately engineered to lure viewers away from the content they want to see: search mesothelioma and you’re presented with legal firms, not information about the disease! Sites like SpyFu can list keywords for you if you’re curious. I’m of the opinion that deliberately manipulating and deceiving those who are trying to get information is completely unethical. Then again, I’m one of those “pie in the sky” internet generation members who can remember not having all the information in the world at my fingertips, so I’m pretty defensive of the internet and freedom of information in general.
So, back to my project. I bought a dating-related URL. My original intention was to create a site based on this formula:
1. Pay people on Amazon mechanical turk to supply me with content. Since I was mostly aiming at the micro-blog format and generic content, I figured it wouldn’t be hard to get content for not much money.
2. Build a farm of fake sites that link to the fake dating idea site.
3. Link the site to a platform like HootSuite and create a Tumblr, Twitter, Facebook, etc. account that would allow for easy automation and linking of all the accounts. One post on one platform(ideally wordpress) would generate an automatic spam storm across all the platforms. Social networks are typically easy hits because people are stupid and trust content on them more than content on the open internet.
4. Aggregate the content, collect a month+ worth of daily posts, link all accounts, schedule posting of content, step away.
5. Sign up for Google adsense and post pay-per-click ads.
It all sounded so rational- the site would look quasi-legit. Lonely bored people will click on anything! In general, any kind of blackhat SEO targets vulnerable groups. Think of the spam you get, and the ads you see online- it’s clear that smart, stable, rational human beings are not the target market of most online ad campaigns. I didn’t have the heart to target fat, lonely, mentally ill, bankrupt, elderly populations. So I thought that churning crappy content out of Mechanical Turk would be the perfect way to maybe provide content that looked legit enough to get a click.
There were a number of problems with this strategy, and it failed completely and totally. Here’s what I learned:
1. The keywords and the target market were mostly worthless. A real challenge with creating blackhat SEO content is choosing target keywords that attract idiots ready-to-click AND choosing a target niche with a balance of high value keywords AND a low competition niche. Not easy!
2. Turns out that people on Mechanical Turk just didn’t want to even write a 100 word post on my topic, even for 25 cents. This was a major stumbling block. It’s harder than I previously thought to create spam content.
3. Click laundering! Most Blackhat SEO sites only serve one purpose- to link to a quasi legit site. The number of sites linking to you increases your page rank. The quality of the sites linking to you has no effect- if it did, then your competitors would build spam sites and link to you. But linking to spam will decrease your page rank. For this reason, spammers create large webs of pointless sites that link up in pyramid fashion to the most profitable site. The pyramid is because the more influential the site that links to you is, the more they increase your page rank. What surprised me here was that the site at the top of the pyramid was often a totally legit, undetectable site!
4. Because of the scale required to generate any real return on blackhat SEO attempts, you’ll have to know a programming or scripting language. I’ve since tinkered with PHP this year, and it really helped explain to me how a lot of sites are created. One of the most popular strategy is to obtain a database of madlib content that pulls from a keyword table that you create. These sites generate posts that pull from your keyword list and fill in the blanks. The trouble with this is: you’ll get blacklisted quickly when google finds the same content in a million places.
In this process of attempting to be a blackhat SEO/blog spammer, I learned a ton about the internet. I figured out why it’s so full of junk content, why google trends and keywords are almost entirely fake, and why google keeps changing their search algorithm at the pace they do. The size and scope of the blackhat SEO operations in effect currently are amazing. Many of the most profitable spammers are running thousands of bogus sites, scripting thousands of bogus clicks, etc. So much so that google trends often detects their traffic as a trend and feeds back into the loop- artificially increasing the value of keywords as other spammers hope to buy them up.
The “legitimate” strategies for SEO are extremely limited in their efficacy and diversity- designing sites to have regular updates, choosing keywords carefully based on publicly available market data, avoiding certain words in titles, etc. These strategies are also rendered useless almost immediately when google makes a change in search algorithm. For this reason, I am extremely suspicious of so-called SEO firms. “Ethical” firms won’t be very effective. Unethical firms could ruin your site’s reputation with one bad move that exposes your site to their spam farm, easily lie to you about the actual traffic coming to your site, etc.
I understand so much more after the experiment, and it was well worth the $120.