Further to my two earlier posts about Search Engine Optimization and authors, I found an interesting article over on “The Book Designer” by Joel Friedlander that has six tips for SEO improvements. Of the six, four of them are a bit unusual (first two and last two) and worthy of consideration.
The first (anchor text) is actually helpful for others to use for you — get them to use some text and the name of your sitedomain in their actual link (i.e. click on “this great site about books at mybooksite.com” rather than click “here”). However, you can use it in your own in-site links too. The second (link juice) is also about how other, more popular sites link to you.
The third and fourth (title tag, first paragraph) are standard fare and I’ve covered in more detail earlier.
The last two (link out, link deep) are about giving links to more information or resources on the topic. You have to be a bit careful with those as you start to look like a link farm if you don’t have a lot of other text to go with it, but some good information.
Check out his article (linked above) for much more detail…
About three weeks ago, I posted my entry about search engine optimization (SEO), with a focus on how it was applicable to writers’ sites. I typed it up, proof-read, hit “Publish”. And less than a day later, good old Google lit up the tech blogs with confirmation of their implementation of tweaks to their Search Engine algorithm. The update, nicknamed Panda, is basically a result of a lot of “bad” things I talked about in my earlier post — spammers figuring out enough of the previous algorithm that they could fake their way to better rankings. Google’s goal is still the same — the most relevant, quality sites should rank first. However, if outsiders figure out the way to game that search, then Google starts to lose relevancy and market share (anyone remember Alta Vista search engine?). So Google monitors what sites do. And then adjust their algorithm to fight the tricksters.
In this case, the analysis available on the web of the changes is pretty extensive. CNET, for example, ran 2000 of the most popular searches before and after the change. And compiled an analysis of which sites’ rankings were affected.
Generally speaking, content farms were reduced in rank. These are sites that basically scrape their content from elsewhere on the web and just replicate it on their site — unlike news aggregators that credit the original source, a lot of scrapers don’t. And often they just mirror their sites in different forms (or scrape each other!) to boost their rankings — essentially trying to get more points by having more content on their site. However, the Panda changes basically take those types of sites and reduce the number of points they get from having multiple, low-quality articles that nobody wants (the sites make money by giving you a little bit of content on a lot of topics and a lot of ads when you visit).
But the tweaks are not perfect — some legitimate sites (i.e. non-scrapers) also got caught by the changes. Like the British Medical Journal, for example. Does this matter to them? Probably not — if someone searches for the BMJ, they’ll find it real fast. On the other hand, if the latest article about dermatitis doesn’t show up first on a general search related to dermatitis, BMJ isn’t going to stay awake at night worrying they slipped in the rankings.
By contrast, if the people who sell ads for scraper sites see their rankings slip, then the number of visitors to their site will slip, the click-through rates for the ads on the pages slip, and their commissions slip. So they need to update their tricks to make sure people are finding their sites, and that WILL keep them awake.
What does it mean for you as a writer? Almost nothing. You still need original content on your site and the right keywords. What MAY affect you though is if you have a lot of material on your site that is just a link to another site, or is a reposting of other material, even videos or news items…your site may start to look to Google’s spider bots like an aggregator, and you could lose relevancy points.
But do you care? Oddly enough, the answer may be no. While that seems counter-intuitive, think about how people get to your site. Most likely they clicked on an URL from another blog, or they searched for your name, or they searched for a phrase like “Books by Jane Author”. As with the British Medical Journal example above, you’re still going to come up pretty high (probably first if your URL is your name).
Rankings are therefore only relevant if you have a lot of other content on your site that you want people to find — like, for example, if you write a lot of blog posts about forensic studies. If you see this as a key marketing hook (i.e. “Come for the knowledge, stay for my books!”), then you should care somewhat. But unless you are the foremost expert, it’s going to be hard to crack those first few pages. For example, if you search for the term “mystery book”, you get a whole lot of sites that sell mystery books. But the rankings don’t produce an individual author’s site until #15 or #16 — http://www.robertburtonrobinson.com — and it looks like one of the reasons he ranks so high is that he has posted a lot of short stories online (i.e. original content). [** Note that I just repeated the search, this time logged into iGoogle so it knows how to “localize” my results, and RBR came up a whopping third this time! Wow!]
For the rest of the Panda algorithm, you may be interested to know that Amazon came in third in CNET’s test, no real change — just further confirming, if you’re selling books, that Amazon is the first stop on a global search engine train, regardless of what Apple, Barnes and Noble, or Smashwords tell you. There is a small caveat to that great Amazon ranking though — Amazon ranks higher on the CNET search as it now sells almost everything from soup to nuts, which means it has more products to match searches. B&N or Smashwords are only going to rank in that global search on items that had to do with books. So not really a fair comparison (as a writer, you don’t really care if people can find TVs at Amazon, just your book, but well, that’s the way the cyber marketplace works. Think of it as “diversification on steroids”. It also improves Amazon’s diversified revenue streams and ensures their longevity, plus their ability to leverage resources for advertising (reportedly $135M last year. And with Amazon announcing this week that they will soon start selling Kindles in Walmart stores (i.e. people won’t have to order it online anymore), expect Kindle and Amazon’s sales to grow even more!). In non-localized findings for “mystery book”, Amazon shows up 4th, GoodReads 6th, Barnes and Nobel not until about 27/28th.
So what else do you need to know about SEO? In my previous post, I neglected two areas of the algorithm.
One area is “registration” and “hosting” data. This essentially means you get more points if you are a registered owner of your domain rather than using a generic hoster like MySpace, Facebook, or on a blog spot. Technically though this is mostly covered by having your own domain name, it just ups the ante a bit to make sure you’re also the registrar for the site (think of it like your ISBN for your books — are you the publisher? or did you publish through Smashwords or Amazon and THEY’RE listed as your publisher? The former gets you more points for your URL.) So if you have your own domain, register it yourself and feel free (where possible) to stuff your registration details with your keywords. Not a huge issue, but take the points if you can; if you’ve registered elsewhere, don’t sweat it.
A second area is more about style. When you write a title for a book or a magazine article, you often try to be a bit quirky or unique — which works in print because when people see it, they already have the context. So an article entitled “What we can expect in 2012” in an article in Business Week is very different from a book with the same title in the Paranormal section of the bookstore or an article with the same title in an issue of Entertainment Weekly. But as a reader you don’t care, it’s easy to tell those three apart. But on the web, particularly in a search engine, the context is often lost — those three titles would be considered equally relevant to a search on that phrase. As a result, SEO gurus refer to a related area as microcontent, and suggest making your page titles more explicit. As a practical example, if your site has pages with books on it for sale, don’t call the page “My books”. Call it “Books by John Doe” instead — give the Search Engine the context of what is on the page if it only looks at that page. You also get more points if the page title STARTS with your keyword rather than just appears in the title — so if your key word is “murder”, then “Murder and Mayhem” gets more points than “Mayhem and Murder”.
Looking at the title for this blog entry, I had originally just written it as “SEO update”, as the context was clear. However, applying the microcontent rule to be more explicit, and to help my rankings in search engines, I’ve changed the title to be “SEO Update on the Panda algorithm and microcontent”. Now, technically I should say just SEO techniques or Panda algorithm or microcontent as the first word. But, microcontent is one aspect of content management, and I’m more comfortable with the title I’ve chosen — you balance editorial concerns against SEO each time you write anything for your site. This just makes that balancing more explicit.
So, I disappeared for most of a year with the arrival of my son. And after I got a whole host of movie reviews up on the old site, DRUPAL started giving me glitches. It never looked quite right, so I opted to just ditch worrying about it.
But in the last few weeks, I’ve thought long and hard about what I really want the site to do and how I want to approach it, and finally came to some basic realizations.
My hoster is fine if I don’t muck around with databases.
If I avoid heavy databases like MYSQL or MSSQL, that takes a whole whack of content management systems out of the equation. Including MediaWiki which I actually managed to get working closer to what I wanted, even though it was using a database. But in the end, it was still too complicated for what I wanted to do, and not worth the overhead.
If I look at “light” CMS packages, there are only a couple that really stand out as worth testing. So I did test them, and right now I really like GPEASY. Admittedly, the documentation is not huge, but my needs are simple. And the guy who created the package and who runs the website forums is REALLY helpful with answering questions.
I managed to start with a FaceBook-like theme, and tweak it almost beyond recognition in colours, etc., and still have 98% of the original code untouched. And best of all, it has a really good online editor that allows me to paste text from a simple editor or just type from scratch, along with lots of little formatting options. Still glitching with the 1.5 line spacing after each paragraph, but I’ll work on that over time.
I also need to ditch a second logo that is showing up, add in separate colours for the background of top and bottom, edit the sidebar, rename the blog, and edit the footer. Then on to Movie Reviews!
If anyone hears of a CMS developer or hoster being shot in the not-too-distant future, tell the cops to check my alibi. I’ve been working for just over two months now on my website, and only now feel like I’m even getting close to what I want. DRUPAL proved unable to handle my needs with regard to the photo gallery, so I went with Gallery 2. It’s fantastic, and works like a charm. I wouldn’t say setup was “simple”, but it was relatively straightforward for an open-source solution, and the wife likes the result, so what more could I ask for?
Well, for starters, a blog. And some static content options for my knowledge portal — something that would allow me to post text in an almost wiki-like layout. I had some success with WordPress and after a couple weeks of good solid design testing, I had a blog up and running. Posted a few things to give it a try, but the menu was driving me batty — never did solve that problem. And then the unstoppable force of website creation ran into an immovable object that is my hoster’s policies.
I have no idea which combination of their default settings was screwing up my server setup, but it locked me out of my own configuration — oh sure, it would ask me for my login and password and if I got it wrong, it told me (so I knew the authentication module was working). But if I entered everything correctly, it just went into a loop that took me back to the login page. Apparently this is a problem with persistent sessions and that my hoster’s variables cause previous sessions not to disappear. Of course, Netfirms Support had no idea what I was talking about, and while I can diagnose the problem by searching lots of tech sites, it is beyond my ken how to explain to them how to fix it. So I went looking for an alternate solution for my other web page needs (and now needing a second blog option too!).
So I tried MediaWiki…my host provider supports an older version of it, and it seems okay. I found it really hard though to wrap my head around the approach — things were still not page-like enough, and figuring out even how to change a logo across similar looking sub-sites was a challenge. Admittedly, the program configuration tells you very clearly that this is NOT the tool to use for a typical website, and it will just frustrate you to no end if you do. Guess what? They were right — I still gave the upfront work a try to see if I wanted it for the knowledge portal at least, but I was just increasing my frustration levels.
So I started experimenting with a site called The CMS Matrix…they’re part of a group called the “Compare Stuff Network — Great data, ugly sites!”. Fantastic site, albeit with a few too many self-serving advertisers rating their own products in a few places. But I went with their top performers and tried e107, Joomla, EZ publishing, Mambo, and a couple of others. Joomla and Mambo were pretty solid options (not surprisingly I like them both as they both started as the same program (Mambo) and then forked into two CMS programs, one still called Mambo and one new one called Joomla). Pretty straightforward user interface. Good testing, still needed to wrap my head around another CMS approach. And then Netfirms intruded again with their persistent sessions. Sigh. I can accept that occasionally my server host might have some challenges with their server config that might cause me some grief, but Netfirms really sucks. They sent me spurious info that was completely useless and unrelated to why I was having persistent sessions (even though caches were clearly disabled! and I even tried enabling them for 3 minute durations hoping that would wipe something, but no!).
Of course, most of these problems are related to trying to use an full-bodied open-source Content Management System (CMS) on what is essentially just a personal site. A personal site on steroids or with delusions of grandeur, but still a personal site. So, since it is just a personal site, I tried a simpler solution using some of the “personal site software” that is out there. You know the ones I mean, cuz they all advertise the same way: “No programming knowledge required! Thousands of templates! Millions of colours”. I checked out a few, but they were all pretty limited. Microsoft Frontpage is pretty powerful, so I considered that for a bit since I already have it, and was what I used to try earlier versions of the site. But I upgraded temporarily to try out the new all-powerful Microsoft Expression. Pretty powerful, and way too complicated. I couldn’t even figure out how it related to Frontpage’s approach. Kind of like going from a Paint Program to Adobe Photoshop Professional Edition. So that option was out. Which left me way too close to my original option of coding the design by hand.
Instead, I seem to have come back to Drupal. Can it still count as a new version of my website if I’m back to near to the approach I took for 2.0, just further along? While Drupal is incredibly powerful, and really WAY more power than I need, it does allow me a bit more control over my interface with the host. Not quite to the “2×4 upside their head” level, but almost to the “hah! here’s a small slingshot to shoot paper wads at you” level. Which doesn’t change anything on my gallery, that will stay separate. But I no longer need MediaWiki or my separate blog site. I can basically do everything I want within Drupal except the gallery.
Don’t get me wrong, it is not a paradise. I’ve spent a good portion of the last week trying out some advanced VIEWS and PANELS and TABS options only to decide most of them are way too complicated for the little sub-projects I want to do. I’ll defer those to version 3.0 perhaps, and instead focus on getting the basic structures and initial texts up on the site. One thing that I haven’t quite figured out yet is how to add images to these posts, but that will come. In the meantime, on to Movie Reviews!
It’s been an interesting six weeks since I started putting some serious time into the website. I’ve played with a bunch of content management systems, and some desktop software packages that design sites out of the box, etc. I even had DRUPAL working pretty well. But there were a few problems with each, and for DRUPAL, it was getting my photo gallery to work.
I mentioned last time it was the bane of my existence to get it up and running, and I reconfigured that sucker ten ways from Sunday. Nada. Don’t get me wrong, I doubt it’s DRUPAL’s fault. It is more that it is a step above my needs and yet a few of the simpler CMS software bundles didn’t handle a few key features.
So, in the end, I’ve gone with a mish-mash of software that my host provides automatically pre-configured out of the box. For those of you familiar with Photo Gallery 2, I’ve got my pictures site going under that. For the most part, I’m pretty happy with it. I’ve found a general theme that I can tweak enough to live with, upgraded some of the modules and plug-ins, and uploaded a batch of photos.
For my general “blog” area, I’m going to go with WordPress for now. It seems to be fairly robust and not TOO complicated (I hope). For the rest of my “static pages”, I think I’m going to go with MediaWiki.
