WordPress.org

Support

Support » Requests and Feedback » Built-in statistics

Built-in statistics

Viewing 15 replies - 1 through 15 (of 24 total)
  • That’s planned as part of my referrer hack. Meanwhile you could just use the webserver log to get statistical data on that…

    I added a “Popular Posts” section on my blog about a week ago. Is that what you’re looking for?

    I think it would be nice to have a blog-aware referrer framework, ie that it tracked hits and associated them with post IDs. It could also track the various RSS hits. This should probably be an official WordPress feature. Let’s start with the database schema:
    –table wp_logs
    log_id
    log_time (timestamp in GMT)
    log_referrer
    log_ip
    log_hostname (would require DNS lookups, perhaps optional? run in batches?)
    log_request (the request URI in full)
    log_post_id (if available)
    log_useragent
    It would be automatically embedded on WordPress pages, but it would be nice to optionally include a file (like include(‘wp-refer.php’); ) from *any* page and log those hits as well.

    Alex King
    Member

    @alexkingorg

    I’ve thought about trying to create a weighted scale for “popularity”. A view on a page with other posts would be worth 1 while a view of the post directly would be worth 5 and a comment would be worth 10 – something like that.

    Hmm, that’s an interesting idea. It would be best to keep it seperate from hits though, since it really doesn’t pinpoint whether a particular post was actually looked at on the 1 point scale. That could be a field in the wp_posts table I suppose.

    Alex King
    Member

    @alexkingorg

    I think that you still need to give some weight to a post being viewed when it is on the home page or similar. I often refer someone (in an IM or similar real-time communication) to a post at the top of someone’s blog by giving just the blog URL (since seeing previous posts often gives them useful context).

    Cena
    Participant

    @cena

    I know that this is not what most people seem to be looking for when they talk about site statistics, but I’d really like to have something like allusion’s (now-defunct) Zeitgeist script, that kept track of things like most commonly used words, total number of words on the site, average number of words per post, etc.

    I like Alex’s idea of weighting the popularity, although I think it should be something that can be configured. I mean, that’s a small integer to change in the database. 🙂
    GFM

    Laughinglizard
    Member

    @laughinglizard

    At this time, I use a javascript refer script which keeps track of hits on a page by page basis. However, as I have my permalinks setup to point to a single post and associated comments, it almost acts like a popularity rating of a particular post. The code is real simple, (from another javascript coder) but uses text files to keep track of visits. Here are some of the items I believe are important to my blog tracking (in no particular order):
    1) Which search engines and which keywords
    2) Which posts get the most attention (possibly a popularity count, no of direct hits, no of hits referred to by the main blog, –more– clicks)
    3) Highest referer count from…
    4) Hits to main blog by day, month and year
    5) Incorporating Alex’s “since last visit hack”, post and comment count
    and finally…
    6) Some way of tracking a hit into a blog. Let me explain myself on this. As we already know how many times an entry has been accessed, I think it would be really interesting and useful to know entry and exit points for a certain visit. Do most people come in through the entry page for the blog or through a search engine into a category, or directly to a post? (this is very true for my weblog tools blog, and I modify the index to be more post centric than main blog centric) I second Matt’s idea of having a wp-refer included in every page. This file could grab HTTP_VARS as and when they appear. In my idea, when a visitor enters the page, an entry for the user is created in the database and is not ‘expired’ till a timeout occurs (which would be triggered by another visit by someone else after the timeout period). This also gives us the power of providing the returning user with useful tracking and follow up information on their next visit, maybe a customized bread crumbs of sorts.
    I know this is convoluted, I have been working on OpenGL all day and my mind is all GLed up, so please excuse my thoughts 🙂 I would be glad to explain myself further or help out . Thanks for reading thus far.
    Peace.

    Sushubh
    Participant

    @sushubh

    Add some browser identification features too. As an Opera junkie, I would like to know that :)…

    http://das.reinvigorate.net has some good stats service. And it’s open source. Maybe something like it can be implemented.

    See the advanced counter over at http://htmlfixit.com which is a $10 item. I am using it on a blog and it gives me the pages hit. Pretty neat stuff for $10.

    Sushubh
    Participant

    @sushubh

    in 10 dollars some host will give u hosting + wabalizer. but thats another thing :)…

    Cena
    Participant

    @cena

    Has anyone done a port of b2stats?

    John
    Member

    @j2000_ca

    $10 is ridiculous, check out my free plug-in. It’s actually an improvement on an existing one but adds many substantial features. We needed a stats plugin for our blog, and like all good programmers didn’t like what we see, and like all great programmers, wrote our own. Stump link: Our homepage. Hope you like the stats!

    Here’s the Link

Viewing 15 replies - 1 through 15 (of 24 total)
  • The topic ‘Built-in statistics’ is closed to new replies.