I’ve grumbled about this before, including in the recent post about a hit for Tracy Hyde pics that, when I search it, brings up no actual result pointing here. Now I also see one for Tracy Hyde photo, same deal. Most of the “search phrases” reported by Awstats are things like attacku3k, pressdjv, changing1gx, holdk6w, etc. Completely strange and bogus. Then again, the referrers are almost completely fake, too: Referrer spam, in hopes you might click them when you look at your stats, or something. Some of them are topical, at least, like one that points to something on how to be assertive.
Why purport to show search strings in the stats when obviously you can’t or won’t? So far this month, through about 90 minutes ago, I have about 437 hits from Google. The rest are trivial. By comparison to supposed hits from other sources like “direct address,” “bookmark,” or “link in e-mail,” search hits are trivial. However, that is based on “pages,” which is a number inflated by spammers or other malicious sources hitting things on the site that might not even be visible to people. It’s what happens when you get a relative monoculture of one convenient CMS such as WordPress. Or even an oligarchical culture of a few such things, rather than a wild west of people writing their own HTML. Then again, the nature of the web is relatively transparent regardless. Back in the day, PCs using a Microsoft OS got viruses or malware, and, as people would say, “Apple doesn’t get viruses.” Aim at the big target.
So really those Google hits are probably actual people less a portion of malicious sources arriving via search, and are some fair portion of the 2317 “unique visitors” so far this month. Yet the stats can’t see what the search strings were and report them? I wouldn’t be surprised if there’s some obfuscation from Google going on, since they are in the business of reporting their own results if you sign your site up with them. Google Analytics, in my experience, is a bit like hitting a gnat with a sledgehammer. Also I haven’t seen actual search strings there, on sites where I have used it, though it’s possible they must be if you look hard enough through the mess that is the interface.
The most useful thing I can discern currently is that an overwhelming amount of my traffic is now from Great Britain, with the next two being a race back and forth between Japan and United States. That’s how it settled out since blocking enough malicious IP addresses to reduce the numbers from the usual suspect countries. What do I post about a lot? Melody, the 1971 movie. Where was it filmed and where did it eventually if not instantly become popular? Great Britain. The actors tend to be from England or elsewhere in the UK. Some of them are actually part of the movie’s fandom. Where was it a runaway hit? Japan. However they’re finding it, that’s gotta be the source of a lot of traffic.
For some reason, this post is particularly popular as a landing page, and has been almost since I posted it, but after that is the category for Melody. Above both of those are the main page, naturally, and the feed. How are that many people reading via RSS? This is not my blogging heyday. Heck, in my blogging heyday, I’d have considered these great stats. I’ve had individual posts avalanched with tens of thousands of hits in years past, but not in a long time. Anyway, after that some of the results include popular ancient posts on an archived site at this domain that was created in Expression Engine and never ported to WordPress. Brave souls, going there when the pages consist of long lists of PHP errors before and after the actual post. The version of PHP on the server outstripped the version supported by EE.
So I can discern a few things mainly by looking at the stats. Melody posts generate some interest. So do some of the music posts. Google sends me a lot of mystery traffic from searches. For all I know, a lot of that goes to the archived site. A weird number of people use the feed. Most of the traffic is nefarious. All I get for comments are spam ones, which I believe these days are generally automated. The mix of pages people visit via HTTPS is substantially different, with the residual politics category and history category being far above Melody, but the total coming that way is dwarfed by plain HTTP.
It’d be fun to see a better report of the search results, but oh well.