It seems everyone has their favorite analytics program or analytics service. After testing and using many of the most popular services, standalone programs and WordPress plugins, my favorite real time analytics program is Mint. The service that comes in second for me is Google Analytics because data retrieval is delayed.
Although using a plugin that caches WordPress is the best overall solution, I can't use any of them with the way I run this blog. I did the research, tried the hacks, and ended up falling flat on my nose. I'll explain what I finally ended up doing after I list the options.
If you want to hide and display stuff in your sidebars, or anywhere on your website, you can do it with hidden divisions and JavaScript. Today, I'd like to pass on some JavaScript code I'm actually using on this blog.
I was asked to beta test a new advertising service called TextAdMarket. This advertising service is very, very new. As with any new service in the beta-testing stage, this one isn't without its share of problems. There are very few members using it right now and fewer is better until the bugs are eliminated from their system. It has only been live for a little over a week.
I wanted to wait until I had the widget on my blog for a month or so before I wrote about it, but I figured I'd forget why I installed the widget ScratchBack in the first place if I waited that long. I want to gauge its effectiveness for traffic, not for gain.
As you can see when you look at the widget in one of my sidebars (it could change positions, so I won't say exactly where it is), I chose the minimalist theme for my widget. The rest were too wide for my sidebars and I don't have a creative bone in my body (or Photoshop), so it was either choose that widget or widen the sidebars. I took the easier route.
Since I have no fondness for slow loading third-party JavaScript widgets, I made sure to make use of Stephen Cronin's IFrameWidgets plugin for WordPress. Except for a couple of XHTML validation errors, it does the job of preventing interference quite nicely.
Tactics
My aim is to take what I receive as ScratchBack tips and reinvest it on the ScratchBack widgets at other blogs. I'm specifically looking for blogs where I think I may draw traffic from. I'm not going to wait to get paid ScratchBacks minimum payment amount before tipping, though, because I have online income from other sources to cover it until I do. I just have to be frugal for the time being.
Of course, if I spot the widget on a blog that I like to read, I'm prepared to drop a tip on a dime. That is, unless the blogger is asking for an outrageous amount.
What took me so long?
I wanted to wait until after the buzz created by the prominent bloggers, and their followers, died down before fooling with it. In my opinion, the effectiveness of ScratchBack tip jars can't truly be gauged until people stop writing about it so much.
I'm prepared to write a follow-up article in 40-60 days, depending on where I am at the time. I'm also prepared to get rid of the widget and stop using it on other blogs to generate traffic if doesn't appear to have much effect. I'll be sure to let you know.
I had to hold off on doing some optimizations until I had a stable connection for at least an hour or so. I think the wait was worth it.
Optimization #1: Removing the last of the social widgets from the footer and any advertising widgets from the sidebars. I'm still making money in other ways, so phooey on sidebar and footer slowdowns.
Optimization #2: Replacing WP-Cache 2 with WP Super Cache. Those who are logged in or leave a comment get the benefit of the WP-Cache 2 that's included with WP Super Cache, but with code optimizations (Donncha picked up where the other developer stopped). Those who are not logged in and don't leave comments (99 percent of the Internet) get the benefit of WP Super Cache, which serves static pages that aren't processed by PHP like WP-Cache 2.
Although there's a noticeable speed improvement, even on my lousy connection, there are more optimizations that need to be done:
I've been trying to incorporate something like Minify for PHP4 (since that's what my host uses) which is based on Minify for PHP5. Although I'm familiar with both PHP and CSS (and even a little JavaScript), I couldn't get it to work for me. My main CSS would stop working completely.
I took a look at Combine, but it appears to be written for Apache 2 and my host uses Apache 1.3. I also took a look at JSmart. It didn't work for me either.
I'm not sure what I need to do to get one of these options working, and I'm all eyes if anyone can help.
Search Engine Rankings: I saw a note somewhere (sorry, but I don't remember where) that Google did away with their supplemental index (Google Hell). I don't know if that, or the recent Google PageRank slapfest, had anything to do with the improvement of this blog's search engine rankings. I do know that, even with a PageRank of 0, I'm getting more traffic than some people with a PageRank of 4 and 5.
It's a known fact that the more often one of your links appears on Google's search results, the faster it climbs to the top of the index (I hesitate to use an acronym like SERPS because I really don't know what it means and what factors are in control). It's not a known fact that there are ways to have your links come up more often, without being searched for. One of those ways is a plugin called AskApache Google 404. I have it installed on both of my blogs. Go ahead; click this link to test it: Google at Untwisted Vortex. You'll see what I mean almost instantly.
I've been working on optimizing my CSS. Before I get into that, a brief history is in order:
Back in May, I made some adjustments to my theme. The first was the Google AdSense ad placements. They're still displayed that way except now I use the Shylock AdSense plugin to display the ads. The second was to increase the font size of the main content from 10 pixels to 11. I noticed that the "Tigarator" options of my theme reset themselves several times since then and that's what prompted me to dump the "Tiga" functions from my theme and start editing the CSS by hand.
After that, I made some more adjustments to my theme. The most significant was making the sidebars load after the middle content area. Then, after I said goodbye to my JavaScript widgets, my blog loaded pretty fast. I tested the speed back then with a meta tag analyzer and it loaded in under 1 second. I tested it today and it still loads in under 1 second (without the widgets).
I boycotted JavaScript widgets the first time I felt the need for speed but have since put a few back. I put them where they wouldn't affect the displaying of the contents, in the footer. I do have one in the bottom of the right sidebar, but it's the last thing to load before the footer.
Last night, I somehow found myself at Daily Blog Tips and read the post about 6 ways to speed up your site. Tip #3 about using Clean CSS caught my eye as well as a comment by Chris from Reflective Reality V2.0 on how to make WordPress quicker by adding a single line to my .htaccess file. I did the .htaccess thing. I wonder if my blog is any snappier now.
I started using the Clean CSS website this morning to see what changes were suggested to my CSS. I made a lot of the changes but stopped because I didn't understand the changes it suggested for the shorthand properties, so I "Googled" for more information and came across an old post at 456 Berea Street that did a very good job at explaining efficient CSS with shorthand properties. Armed with that information, I now feel confident enough to complete my CSS optimization project, which I hope to finish today or tomorrow.
I spent the greater part of yesterday, and part of today, experimenting with things like output caching for JavaScript widgets, tweaking various parts of my theme, testing program flow with PHP, and testing effects using CSS. This probably sounds really mundane to the average person, but it's the stuff I learn and like to pass on when I discover something new and not well known.
I found out that it really isn't worth it to try and cache JavaScript widgets because of their dynamic properties. I just wish the designers would offer image-based (like MyBlogLog) widgets that defeat the hang-ups caused by slow responding scripts.
PHP has so many different ways to control program flow, it's mind-boggling. You can use conditional "if" statements and case statements, just to name two of them. After reading many, many tutorials, I found that many programmers forget about one aspect: Fall through. An "if/else" statement can be reduced to just an "if" when there's only one condition to test for. It will fall through to the next part of the program if the condition evaluates to false.
CSS (custom style sheet) properties really can do everything that tables, rows and data cells used to be used for. No one noticed when I removed the tables from my header and started using only CSS properties. I didn't write the table data as part of the theme (the theme designer did), but had added an additional cell. When I replaced the standard search with the Google search, that's when I switched to CSS properties for the design. Yesterday, I tried to use CSS to adjust the vertical alignment of the FeedBurner chicklets, but they just wouldn't cooperate. So I did something different and scrapped any attempts at styling it. I did manage to get the "single page footer" altered though, and added RSS icons down there.
I never realize how time-consuming these things are until after I've spent hours doing them.
After I finished Saying Goodbye To My JavaScript Widgets, I knew that my blog would load faster. How much faster, I didn't know because I didn't have a tool of any kind to test it with.
Sometime yesterday, I stumbled upon (no pun intended) a blog where a tool was listed in the comments. I don't remember which blog it was or who the commenter was, or I would definitely link to both of them. The tool is a hosted one, at widxl.com – Internet Solutions, called "Meta Tag Analyzer". There are some other tools at their site you might want to check out while you're there.
What I discovered by running the tool is that my site loads via broadband in a second. 1 second! I tested some of my favorite blogs and found that most of them loaded in 2 seconds or more. That's the difference between having the JavaScript widgets and not having them.
There's something else I noticed since dumping all the JavaScript widgets. There's not as much disparity between Google Analytics and Clicky as there used to be. Clicky only registers a visit when the page completely loads. Google Analytics, because of the plugin I use, registers it much earlier. My guess is that visitors with slower connections were clicking on to other sites before the page they were on even finished loading.
As of right now, I'm boycotting any third-party widgets that use JavaScript only and I recommend anyone who's interested in speeding up their websites to do the same.
Although I like the widgets in my right sidebar, I will be removing forever those that do not have a non-JavaScript solution. I don't like slow page loads on other blogs, so why would I put up with it on my own?
Some of the widgets have non-JavaScript solutions and at least one has a different widget without JavaScript. The ones that have JavaScript only options… all I can say is goodbye to that service.
After editing my theme pages to make the center content load before the sidebars and shoving those widgets to the bottom of the right sidebar, they still hang sometimes and prevent the footer from loading completely. This is unacceptable.
I can't do it all right now, but if you come back in 24 hours it should all be done.
Update: Done!
RSS Feeds: