I was doing some reading on one of my favorite forums the other day, and a discussion had broken out about Google’s sandbox, whether or not it existed, and what its actual effects might be. I identified three competing theories about the Sandbox-phenomenon, but I only really agreed with one of them.
Beyond just the issue of the sandbox (another post in itself), it got me thinking about where misinformation comes from in the SEO world. How do people end up with such wild information about everything related to SEO?
Telephone Game
One of the most obvious sources of misinformation in any industry is mis-translation, either through laziness, error, or lack of comprehension.
For instance if, in a client meeting, I made the statement: “Your Google PageRank will only change roughly once every three months, and is largely dependent on the rank of other sites linking to you,” someone at the meeting may misunderstand what I said.
Later, they may repeat to someone else that “Where you rank on Google pages only changes once every three months.” Although PageRank and search-ranking are two totally different concepts, someone may now be infected with the misinformation that one has the properties of the other.
Extrapolation of Results
This might be more commonly known as the “knowing enough to be dangerous” effect. Generally, this type of misinformation comes from someone who has done something of little significance, and applies to some greater situation. I believe this to be especially common in the web marketing world because most people keep their exact niches/industries a secret in order to prevent competition.
For example, a web designer in South Dakota designs a web page for a cosmetic surgeon in the local area. He reads a bit about SEO, and so he optimizes the site, maybe gets a few links, and helps the doctor submit his site to Google Local. Voila! Withing a couple weeks of the site being up it ranks for “south dakota plastic surgeon” and bunch of other great terms.
Meanwhile, in San Diego, an SEO firm is feverishly building links for their new, local cosmetic surgery client. Despite getting awesome links from authority sites, perfecting their on-page optimization, and having a rock-solid web server, they can’t seem to get to the top of the page.
The designer and a member of the SEO firm meet on a forum after the desperate SEO posts a question: “We’ve done x, y and z for this local cosmetic surgeon’s site and can’t get it to the top. Can anyone think of something we haven’t?”
The South Dakota designer chimes in, “I just ranked a similar site in my local area for basically the same terms. I’d recommend taking a look at…”
“Thanks,” says the SEO, “we’ve done all that. Any other ideas?”
“You’re probably penalized,” replies the designer. “You should resubmit and look into the different penalties.”
Notice that the two never compared market sizes or competition. There was simply an extrapolation of “well, he did it, we must be doing something wrong.”
I can tell you right now that the market for cosmetic surgeons in San Diego is at least three times as competitive as the one in South Dakota, and probably more like 5-10x. Months later, that same SEO might try to “help” someone else by telling them about the penalty they received.
Compounding/Confounding Variables
I’ve seen this written both ways, and I believe they are equally correct, although they describe different parts of the same phenomenon.
“Compounding” refers to the multiplied affects of two or more variables affecting the outcome in their own way (could be the same, in which case your result is increased, or different, in which case it is sandbagged).
“Confounding” pretty much refers to what happens when you look at the results. In the case where two variables affect the outcome in equal, but opposite directions you would see no result. Totally confounding.
Given the massive number of variables in any SEO situation (from server setup, to domain entries, to site architecture, to content structure, to off-page factors, and search engine algorithm changes) it is easy to assign some outcome to one variable, when it was actually caused by another.
For example, take the case of the PPC manager who writes some new ad copy. She starts with her original ad:

And then writes some new copy:

“Interesting,” she thinks to herself, “that new ad copy makes a very definitive arrow pointing right, I wonder what that will do to click-through rates.”
She has just introduced a compounding variable. Later, when the new ad outperforms her old ad she may attribute the success to the geometric properties. But look at the ad, it’s just better all around.
Isolating those types of variables is difficult, but can be done. Besides, is it more likely that the ad improved because it uses both ‘clients’ and ‘customers’ and has more pleasing flow, or because the copy points to the right? I would say the former, but we could now test for the latter as well.
When receiving advice or information on web marketing, I would urge you to consider the source and parse for any of the above. Too many mistakes are made because the source of information is unreliable.
Tags for This Post:
game one,
lack of comprehension,
cosmetic surgeon,
search ranking,
telephone game