Published on February 2nd, 2013 | by Derek Slater0
Improving search ranking for editorial content: A real-world test
This article was originally published December 2010 on emediavitals.com.
In my previous column way back in July, we discussed two absolutely critical points in Search Engine Optimization:
- Only the first results page matters, and
- That means the second page is your low-hanging fruit.
The reason for point one is that, for any particular search, the items on the first results page get around 85% of the clicks. Content items ranking lower get only a fraction of the traffic for that search. Page one gets steak; everybody else gets scraps.
But that also leads us again to why I love discovering that an article on my site ranks on page two. A page-twoer has already demonstrated that Google likes it. It just needs a little additional push to start bringing new readers to your site. Enough of a nudge up onto the first page.
In the aforementioned column, I gave myself a public challenge, listing five CSO articles that appeared on page two for five specific searches. In this column, let’s look at the results of my efforts (so far) to nudge these specific articles onto page one. We’ll also look at three additional examples that help round out the view of possible outcomes, good and bad.
The five-example challenge: Grading the results
Our work on these articles will demonstrate some of the weapons in your arsenal, as well as some of the obstacles you may face, as you try to improve your own search SERP standings.
SAS 70, which was #11 for the Google search ‘SAS 70’ when I started this exercise. SAS 70, incidentally, is an auditing standard relevant to security folks.
TACTICS: Included a link in a new article about the standard that will eventually replace SAS 70. Also, to state the obvious, I linked to this and each of the other examples from that previous column here on emediavitals. Every inbound link helps!
OUTCOME: It’s been as high as #6 and is now #12 – one slot worse than where we started.
GRADE: As we say on the Interwebs: FAIL! Not a spectacular failure, though – more on that later in this column. This SAS 70 article is still perched near the top of page two. So close and yet so far.
This exhaustive article about encryption ranked #12 for ‘end to end encryption’.
TACTICS: Linked to this article from several other old ones on the site; added the keyword to one subhead (using h2 tag).
OUTCOME: Now ranks #4.
Social engineering: The basics ranked #14 for ‘social engineering’.
TACTICS: For the most part, links in new stories. We regularly cover variations on social engineering, so this was natural and easy.
OUTCOME: Now ranks #9.
GRADE: Very good. Last month ‘social engineering’ was one of the top dozen phrases bringing searchers to CSOonline. Of course, we are hoping for further progress.
How to compare and use legal hold software ranked #16 for ‘legal hold’ and #18 for ‘legal hold software’.
TACTICS: Rewrote the lead and featured the phrase more consistently throughout the article. Added inbound links, including in an Essential Reading box.
OUTCOME: This one has been flaky. For some months, a duplicate version that ran on a sister site, NetworkWorld, ranked higher than our original version. In spite of correct use of canonical tags on both sites. Now (to my surprise, since I had given up) I discover the CSO version on the first SERPs – #9 for ‘legal hold’ and #2 for ‘legal hold software’! These terms do not appear to have a huge volume of searchers, but visitors arriving via that search have excellent time-on-site and pages-per-visit numbers. So the article’s overall reader service is great.
GRADE: Very good. If this were a more common and competitive search phrase, I’d grade the same results as excellent.
DDoS attack survival guide ranked #18 for ‘DDOS attacks’.
TACTICS: Of the five examples given here, this one got the least attention, although I did create some internal links on our site and on a prominent landing page. As with example 4, there was a period where a syndicated copy elsewhere was faring better on Google, so my motivation was low.
OUTCOME: Slid gradually down the SERPs. Out of contention altogether.
So what’s the bottom line on this exercise? Three successes and two failures. I’ll take that batting average.
Two observations about this process: One, these tactics aren’t tricks. They’re about clarity of labeling and about highlighting your most valuable content. Two, you can’t predict the outcomes exactly. You can only do smart things and know that in aggregate you’ll see positive results.
Three extras: Best- and (almost) worst-case outcomes
Let me throw in three other examples that weren’t in the scope of the previous article.
One: ‘internal investigations’. After building links to this article on my own site (and via syndication deals, on a few sister sites) for several months, I gave up. Then later it abruptly shot all the way to #1, where it sits today. I assumed this must have been a result of inbound links from some other site or three, but I haven’t found them. Anyway, sometimes the pot doesn’t boil until you stop looking at it. This is an example of “lots of effort, which may or may not have had any impact on the outcome”.
Two: ‘money laundering techniques’. The story wasn’t presented with that specific keyword phrase in mind, but that’s how searchers started finding it. So we tweaked the headline to match that search more exactly. Less than a week later the story rose from #9 to #4 for that search. This is an example of “planting tomatoes on Tuesday, eating BLT on Wednesday.” Most of the time you don’t see results that quickly. Certainly it’s easier on low-volume searches such as this one.
Three – here’s the glimpse of spectacular failure that I promised earlier: ‘retail security’. I worked at this quite a while and my reward was a sudden DROP down to the fifth or sixth SERP. Now when you do something to help your content and you end up hurting it, that’s pretty bad, and that’s what I feared I had done. (Not that page six versus page three makes much difference traffic-wise, which is why I am inclined to take the occasional risk.) Happily it bounced back up onto page 3 after another week or so – but this would be an example of “oops!”.
Why would I care so much about this particular retail security article? Here are a couple of vital stats for visitors who find it via search: More than 7 pages per visit. 11 minutes on site. Bounce rate 30% lower than the overall site average.
So I’ll leave it alone for a while, but I will keep my eye on where it ranks, and look for useful places to link to it.