The Best Promotion: Let Google Optimization Do It

There are dozens of ways to promote at Associated Content: befriending people, reciprocal PVs, leaving links like seeds in the forums and in comments. Or you can move outside of AC: use Digg and other social networking sites, chat on forums, use your own popular blog, or promote newsletters and your RSS feed.

But after reviewing my articles and looking at what's performing really well and what's not, I have only one conclusion: there is absolutely no replacement for Google.

Using Google for Promotion

Google right now has 50.8% of the marketshare of people running searches right now -- slightly more than half. In addition, several other search sites use Google data to run their own searches, which means if you optimize your articles for Google searches, you'll be optimizing for well over half of all web searches.

The number one way to optimize your article for Google searches is through keyword placement. Keywords are simply the words and phrases you would use to find an article like yours on Google. Placement ideally should include your article's:

- URL
- Title
- Subheader
- All headings in your article
- Twice in the first paragraph
- Once in the last paragraph
- Every so often throughout the text

AC automatically transforms the first several words of your title into the URL, so you should ensure your keyword gets in those first few words. They also look specifically for no more than 4% keywording in your article. Today's Google is also looking for Latent Semantic Indexing, words you use that are closely related to your keyword; these related words may well count for more than your keyword itself.

Google's Inbound Link System

But it's not all keywords. When Google started, everyone was using keywords, and a lot of opportunistic marketers were using "keyword stuffing" techniques -- the creation of ultimately unreadable articles that were just there to hold keywords and drive traffic to a site. So Google's creators used a different technique to find articles. Instead of depending on keywords, they depended on inbound links. More, those inbound links were ranked by their own popularity.

Here's how it worked: Google examined your article for keywords to determine what kinds of searches it would show up in. Articles were given numerical values that corresponded with whether the keywording seemed to align properly with a topic. (This helped cut out the stuffers when Google eliminated preposterously high keywording percentages and keyword tags.)

Then it examined the sites that had links to your article. For instance, let's say your article was linked to by two moderately-popular blogs, your mom's website that hardly anyone goes to, and one beautiful gem of a link from Vibe. These sites would have rankings: let's say mom = 0, the blogs are 4 and 5, and Vibe is a 9. Your mom's website would give you no boost at all. The two links from blogs would give you a fair boost. But the link from Vibe would give you a serious push upward. Google puts the numbers from each of these links together, runs them through a secret algorithm, and uses that result combined with your keywording to determine approximately where in that search your article would place.

How My Articles Do With Google Promotion

The rule is simple: the higher you place in a Google search, the more likely you will get those page hits, and get them consistently. So how much difference can it make?

Considerable. I went through my articles that have the most page hits, focusing on the ones with over a thousand. I selected a keyword phrase that made sense for the article, then searched it. For instance, I used Homemade Kentucky Derby Pie for my article, Homemade Kentucky Derby Pie. Then I looked in the upper right corner for the number of returns I got. If I didn't have over a million returns, I shortened my phrase: Kentucky Derby Pie, and finally Derby Pie. When I found a phrase that returned 1 million + pages, that was the article whose Google placement I looked at.

What I found out: when I used the obvious keyword phrase each one was centered around, these high PV articles were invariably on the first page of Google results. Most were "above the fold" - i.e., visible onscreen without scrolling down. Several were the number-one page returned for that search. (In one case, the number of returns were over 5 million, and the article was about halfway down on the third page -- but it had mind-boggling pageviews for an article I hadn't even promoted.)

In every single case, the pageviews were consistently high over time. Each of these well-keyworded articles works well as evergreen content.

To be honest, I was stunned by my own results. These articles, numbering about a tenth of my total, brought in more than 75% of my total pageviews.

Also, I did nothing whatsoever to gain inbound links for these articles, and I haven't found any serendipitous ones. That means that the placements I've achieved are solely through keyword placement, which I don't really kill myself to get - just focus on headers and getting at least one in the first paragraph and one in the last. Nothing really fancy.

Impact On Up-Front Payments

There was no discernible difference between my up-front payments for the high-pageview articles compared to the not-so-high ones. That tells me that whatever AC is basing up-front payments on, it isn't some special algorithm they use to parse out your keywords and figure out where your article will place. You are on your own with that.

I imagine what they do is examine your past performance, and then base your offers on what your performance in the past has been.

Conclusion: What I'll Do Next With Google Promotion

My next logical step is to see what I can do to improve placement on pages by using inbound links. There are dozens of tricks you can use to do this: writing articles that point back to yours, posting your RSS feed, using your blog (or other peoples' blogs) to highlight your own articles, registering on Digg and all those other social networking sites, etc. I'm not sure yet what works best with these, but I'm sure some experimentation will tell me.