You've seen all of the hard work that I've been doing lately on the site (or maybe not), and you may be wondering why I thought this was so important that I chose to spend all of my spare time working on it for the last 5 days or so. The answer? SEO.
According to stats, around 69% of traffic to the site comes from search engines, with most of that coming from Google and Yahoo! (we get some from Bing, too). If we average around 2000 unique daily visitors, that works out to around 1400 visits a day that are owed to Google, while most of the rest come from either GameRankings, a few forums or people just tying the address into their window. So, the goal here is to turn 1400 into something more like... maybe several times that, in time.
SEO is the answer, because obviously what we want to do is come in near the top of relevant search results. In the past, there's been some temptation to just toss words like 'sex' and 'lesbian' and 'naked' and stuff into meta keywords and call it a day, but even when that works you're just getting a lot of horny visitors who will leave in disgust the minute they arrive. The key here is to bring traffic not at all costs, but because we have something to offer that we want them to see.
I'm no SEO expert, but I have done a LOT of reading on the subect (including PDF literature provided by no less than Google) and I've learned that in general, search engines are a bit like a warehouse full of glass jars. Any page that you have in that warehouse can be represented by a glass jar, and when a user performs a search he is basically seeking out a glass jar. Google (or other engines) will say "This jar has the most pennies in it, then this one..." and so forth. In theory, then, search results are just a sorting method to let users know which jar is likely to have the most pennies. So the trick is to do what you can on your end to make sure that your page/jar has the most pennies in the eyes of the search engine (and in reality, if your site is a good one, like HonestGamers).
Some webmasters will try to game the system. They will get a bunch of inbound links from link exchange services to put pennies into their jar, or they will spam other sites' blogs or they'll stuff their pages with keywords that might not even relate directly to content. I'm not a fan of those approaches, so what I wanted to do was to make sure that each and every one of our pages put our best foot forward while referencing content that users could genuinely expect to find on each page. This will ensure that when someone clicks a link, he finds what he's looking for and is happy to stay on the site and look around for a bit. This in turn will hopefully lead to community growth.
The changes I made recently help with that. Our page titles already are quite optimized to establish brand recognition (by putting our site name at the front of each page title, which costs a little in ranking terms but will make our site's name familiar to people who frequently perform searches) and to place only important keywords. Our meta keywords and descriptions are also good, plus pages are optimized so that key terms are placed within headings. The only thing left to do was to truly optimize URLs. For an example of what I mean, just search for something like HonestGamers Chronicles Mystery Curse Ancient Temple. Notice how those key words are highlighted all throughout the results? This includes the page title, the description and finally, the URL.
This change was important, and there are other benefits. Now that our URLs are shorter, more people are likely to link them without truncating them. And when they link, by default they'll be including the important phrases within the link, which improves the value of each link because it establishes relevance. Nice, eh?
I'm happy with the changes for the reasons stated above, but I made one final change. While updating things, I noticed that many of our pages have several versions floating around on Google. To go back to the 'jars' analogy, this is like collecting pennies in five jars instead of one. There's only a limited amount of traffic, so every jar would take longer to fill. Since only the fullest jars mean jack squat when it comes to search results, duplicate pages were going to bury us. Now that will be fixed.
All of the chagnes will take months to spread across the Internet, and it'll be awhile before the old pages vanish. Once all of that takes place, which could take 3 to 5 months, I'm confident that my work over this past week will really bring in the traffic. In the meantime, it's not a bad idea to remind your friends that we exist. Nothing says that we can't get a jump on building up our community of reviewers, right?
|Most recent blog posts from Jason Venter...|
|radicaldreamer - December 11, 2009 (07:59 AM)
Nice work Jason. I appreciate and admire your industriousness in working on this site both as your personal project and our community space.
|aschultz - December 11, 2009 (03:29 PM)
Cool and interesting. It's neat to read how you go about maximizing things like search hits and so forth. Having had a geocities webpage once and never had the motivation to kick it up a notch, there's a lot of good technical stuff I wondered about but never had the time to look through & so I'm glad someone else is taking the time to explain it.
|wolfqueen001 - December 11, 2009 (05:23 PM)
Yeah. This stuff sounds like it could produce good results, which is nice. I do have one question about these "duplicate pages", though. Will removing these negatively affect anything that had been archived before? (i.e. Sportsman's tourney archives). I don't really think so, but I'm not sure how these duplicate pages are even created or what their original purpose was or anything about it, really.
|honestgamer - December 11, 2009 (08:13 PM)
Sportsman's archives shouldn't be affected if he has all of the content saved on his server. I'm not sure how he even archives such things, honestly. Those pages also won't interfere with any of these new efforts because they have entirely different URLs.
The duplicate pages were actually one page with different URLs. If you look on Google even now, you'll notice that some pages come up with several different URLs that all take you to the same place. As I changed URLs in the past, I didn't set up redirects to let the search engines know that anything had changed. So then the site would go through and treat the revised URLs as a new page, rather than saying "Oh, I bet systems/games.php?console_id=3&game_id=4 is the same age as systems/games.php?game_id=4" (which would have been the case). Google also would treat "http://honestgamers.com/" and "http://www.honestgamers.com/" as different pages, and when we had "http://www.honestgamers.co.uk/" running, that made things even worse. Much worse, which is why I pulled my head out long enough to remove that. In the end, it wouldn't surprise me to learn that Google has as many as 7 or 8 versions of some of our older pages, which is really just a ridiculously bad thing that is helping no one.
I find all of this quite fascinating and very important to the site's future. I used to wonder in the past why the awesome changes taht I was making had such minor impact on traffic (when clearly the site has the content that it needs to bring in a lot of people), so finally finding the reason felt great. I do expect the next few months--once Google catches up to all of the changes--to be very interesting. If we see a sudden boom in traffic here in a few months, we'll know why.
And yes, Google really does take a few months to catch on. Even if new pages are being indexed right now--and they are--the results tend to appear in the search engine in quarterly batches, give or take a few months. That's another thing that has always made it difficult in the past to tell how much good any changes to code did. Ah, the life of a webmaster!
The good news is that due to the democratic nature of the Internet, all of our competition--even IGN and GameSpot--have to overcome these same issues. They, uh... they seem to be doing just fine with that.
|bloomer - December 12, 2009 (02:03 AM)
I'd like to further optimise the google presence of my own websites, but wIthout full automation of page headings like you have over this site, it's a pain. I don't have a crazy number of pages, but I have to change keyword/title info manually.
Ideally I'd like a search for my name (my real name, not my artist name) to bring up my music sites in the top 10. But my music profile isn't big enough for that to happen.
Still, what I did recently was just to go through my various sites and reinforce the links between them, except where I thought that was being cheesy or annoying. If you search 'Wade Clarke' on google, I have 3 of the first page google results, including 2 and 3, plus 2 image results. Find X and one of my Aeriae press shots. Somedays I have 4 results... my personal homepage seems to wiggle in and out of the front page.
I consider this to be a pretty good overall result, considering there are tons of Wade Clarkes in the world, and also that the only guy who is ahead of me (at #1) actually put his name in his URL. But I could probably do better with more aggression :)
|zippdementia - December 12, 2009 (09:25 AM)
Excellent. I'm glad to be back just in time for optimization.
|darkstarripclaw - December 13, 2009 (09:14 PM)
How are urls 'shorter' when before I could link to a review or guide with the site name, /review or /faq and then a four or five digit ID?
Other than that it sounds fine.