A (swiftly debunked) rumour was going around this week that Google was taking a look at email newsletter statistics (or at least what it could infer from Gmail user behaviour) and applying that data to SEO rankings.  The theory was along the lines that if nobody was opening or clicking through your email newsletters, they probably suck, which means your site must suck too, and therefore should be seen as less of an authority on the internet, and ranked accordingly in the search results.

Though it’s not true, it’s pretty trivial to send newsletters from another domain if you’re worried about it being a factor in the future, though as with most SEO advice, working on providing valuable content in every form of communication is probably a better use of your time.

What’s interesting to me is how much Google can influence website publisher’s behaviours, and maybe rightly so, since search traffic is a huge source for many, though I’d caution you to make sure you’re also looking at social, paid, and instructional (telling people to “go here”) traffic sources, among others.

Let me put this another way: in the next 3 years, I fully expect Google to influence fashion through SEO fear.  Here’s how:

A rumour will go out that Google’s commissioned a study saying that people who wear red pants are 39.67% less trustworthy. Using their network of satellite photos and Street View images, combined with some trivial facial recognition algorithms tied to profile pictures already on the web (because they’ll continue to fail at social,) Google will be able to identify owners of websites, detect what colour pants they’re wearing, and change search engine rankings accordingly.

In their defence, they’ll tell people that they advise you to wear pants that are valuable to others, and basically follow common pants sense.  This will start as an underground rumour, then it’ll break out, then Matt Cutts will debunk it, and this very blog post will have some role, though I’m not sure what (for the record, my pants at the moment are charcoal grey with pinstripes.)

And regardless of the official statement, behaviour, along with pants, will change, all in pursuit of web traffic.

Photo by R/DV/RS

So here’s what getting to the number two search result for a certain keyword can do to a site:

Prior to the spike I think that site was ranking on Google on page 2 or 3, which I’m told doesn’t get you very far, but there are regional customizations to the search results that make it hard to figure things out these days.

Now, before we get to the keyword in question, here’s a breakdown by time of day for one of those days:

The peak is at 5PM Eastern.  80% of the  search traffic that day was from the USA, so let’s assume an even distribution across the time zones and say around 3:30 would be the busiest part of the day for search traffic.  On a Monday.

And there are all kinds of assumptions and generalizations going on here.  This chart is for traffic, and doesn’t really tell anything about search volumes for keywords, since the site title and description are what’ll drive clicks, and I doubt it’s a linear function.  As we’ll see in a minute, for this particular keyword, the site isn’t a great match for the average searcher, but the point is that for the majority of this traffic, someone had to initiate a search.

The thing of it is, the keyword I happened to be ranking for was porn.

Yep, it’s a popular term.  No, I don’t run a porn site.  For some reason back in November (and possibly around Christmas) Google was trying out a new algorithm (or just dancing) and was ranking sites about “(insert word here) porn” better than sites about actual porn (I think Chart Porn was the number one result.)

What really got to me was the time of day thing.  I always figured searches for porn would peak around, oh, just after work, then maybe the 10PM to midnight window, and then I guess first thing in the morning, you know, to stay motivated during the day???  Adjusted to a time zone, anyway.

But here I’m seeing a huge gathering amidst the after lunch crowd.  Which tells me as much about the economy as anything: are there that many unemployed people, or are the people left with jobs really that unproductive?

Learned from an email from the StomperNet guys: having a keyword rich domain name doesn’t make much of a difference in and of itself, but if people are linking to you in a link directory kind of scenario, there’s a good chance that the link text is just going to be your URL, which does matter.  Aha.

Of course, it helps when you’re trying to get links if you can direct the link text to the keywords you’re trying to optimize for (it’ll at least make up a bit for all the “click here” link text that clutters the link.)