Why the SEOmoz SEO Quiz Is [Almost, in a Philosophical Sense, Somewhat] Completely Wrong
SEOmoz just posted a quiz to test your mad SEO skilz.
It’s [almost, er, etc.] completely wrong.
And I say that with the utmost affection for a certain yellow-shoed person who’s coming to my house later for Buffy night. I get the idea. Basic SEO principles exist that anyone operating in the online space should know. The trouble is that SEO is art as much as science, and to top it all off (like those little sprinkles on top of cupcakes), things are always changing. Even the basics change.
Sure, you should always create a usable site, with compelling and unique content, and you should do keyword research and use those keywords in places like title tags and headings, and if you block your entire site with robots.txt, it’s unlikely to ever get indexed. And try to get a few links if you can. Some things don’t change.
It’s impossible to know everything about SEO because some things aren’t hard and fast rules. You also have to pay attention the trends, see what works, keep evolving your tactics. And anyway, I would bet that there’s no one person who even knows 100% about how a search engine handles each step of the crawling/indexing/ranking process. Even those within the search engines don’t know all the parts to to the process. Yes, even him. Go ahead, ask him. I’ll wait. He’ll tell you the same.
But why is this quiz so wrong? Too many of the questions go beyond the science and into the art of SEO, where things are fluid and changing and open to interpretation. The quiz is more accurately a test of how much of your personal SEO philosophy matches that of SEOmoz. Like a match.com compatibility test. You and SEOmoz, sitting in a tree. Romantically optimizing title tags.
For instance, which is less useful and accurate: Alexa traffic toolbar data or Google toolbar PageRank data? I posit that it’s not a cut and dried answer. (Cut and dried? Where did that come from anyway?)
What’s the best way to avoid internal duplicate content? Block redundant versions from crawlers? Maybe. Unless those versions have a lot of links, in which case you might be losing a lot of link credit.
Does internal linking help raise the PageRank of internal pages? Well, maybe. Or maybe it just helps the discoverability of those pages and the PageRank comes naturally by virtue of being part of the site. It’s impossible to know because Google doesn’t publish PageRank numbers. You can’t really experiment to find out. (And now we can debate about whether you can use the toolbar PageRank numbers to experiment with something like this. And then once we’re done debating, things will probably change anyway.)
When should you leave the meta description of a page? Er, never? OK, I suppose if you are unable to create unique ones for every page, it doesn’t make sense to have the same one everywhere. But this whole idea of leaving it off when the page is targeted at long tail queries makes no sense. Or at least, it’s definitely more of an up for debate topic than a fundamental truth of SEO. How many disparate long tail terms are you targeting on that one page anyway? If you can’t come up with a description that fits the entire page, shouldn’t you just create multiple pages anyway for each completely separate topic? And if you do shove a bunch a bunch of unrelated stuff together on one page and the meta description isn’t relevant for queries that return that page, search engines are going to use relevant text from the content rather than the meta description anyway. Existence of a meta description tag doesn’t prevent other content to be used as the snippet instead when that would be more useful.
And then there’s the geeky side of me that says: HTTP server response that indicates a file no longer exists or isn’t working? For one thing, the same HTTP code isn’t used for both. For another thing, I know the answer they’re going for is 404, but that’s not right in either case. 404 means doesn’t exist. It’s 410 that means no longer exists. And “isn’t working” could mean anything! Is the server not working? Was the request bad? I wouldn’t be so particular, except I did sort of write the book, er help topic about it. Also, did I mention that I’m really geeky? I even have a favorite HTTP status code.
And then we get to the question that made me decide to do this blog post.
When launching a large amount of new content on a new or existing domain, the best way to release content for crawling and indexing is:
I got it wrong.
And the correct answer, when explaining why I was wrong…
I answered All at once – to allow the spiders to find the content and digest it in their preferred format. But apparently the quoted me said Over time, in multiple smaller sized chunks.
But here’s the thing. That quote was about using a 301 to move existing content. I did indeed recommend doing that slowly for a number of reasons. You want to make sure you’ve impelemented the 301 correctly. The moving pages might experience a short-term dip in indexing in ranking and you might not want that to happen to your entire site at once. You want to be able to pinpoint problems and not troubleshoot your entire site architecture.
That quote was not at all about creating all new content. Add as much of that as you want! All at once! Well, unless you’re creating millions of pages of autogenerated content! That might look suspicious! (Just saying.)
However, I admit that at least one of the questions has a black and white answer. There is no apparent search engine rankings benefit to having a keyword-matched domain name (eg www.example.com for keyword “example”). FALSE. Otherwise, I’d get traffic from searches other than porn.
Edit: My title, of course, is meant to be hyperbolic and over the top and well, easily identifiable as purposely linkbaity (er, and funny). I have a strange sense of humor. I also like potato chips in sandwiches. We all have our oddities. However, I have gone back and slightly edited the title in case it comes across as unduly harsh. I’m not saying all the answers are wrong; It’s that I feel the quiz tends to ask the wrong kinds of questions — questions that have debatable rather than black and white answers. There are more fundamental concepts in SEO that don’t have these shades of gray. One thing that’s great about the particular questions asked though is that they’ve started a dialogue about those topics. And not only about the topics in the quiz, but about the larger topic of search engine optimization as an art, trial and error, and ever-evolving. So in that sense, perhaps the quiz was [almost] completely right.