Note: feel free to ping me or leave a comment if you are having major issues!

Yesterday the event tracking on a site was not working, and someone thought it might be that I had missed a call to the ga.js file or some extra Javascript. My first thought was “This is why I hate Javascript.” That thought recurred throughout my investigation process, but I decided to save some people time and effort by posting a few things you could check before giving up and crying. My particular example was an onClick event, but the same things could apply to other situations.

1. Check the JavaScript call. So you must understand that Javascript (which I shall now write as JS) can be the best and worst scripting language. Very powerful and can do cool things like animation (and your analytics tracking), but also a pain in that you must be very careful with syntax and punctuation. One easy way to see if your JS is creating errors is by installing the Web Developer’s Toolbar in Firefox, it will show you when there are JS errors and a myriad of other useful things.
   

   Example: onclick=”pageTracker._trackEvent(“Foo”,”Bar”,”JS”);” This will be wrong because you have double quotation marks instead of apostrophes within the parentheses. So the page will read it as “pageTracker.trackEvent(” which will not do anything. Fix it like this: onclick=”pageTracker._trackEvent(‘Foo’,'Bar’,'JS’);”.

2. Is the variable actually pageTracker? In most cases, if you’re not changing the Google Analytics code that is on every page, it should be pageTracker. If the variable is named anything else, like movieTracker or pageCounter, then you need to fix the call in the onClick and change it to: onclick=”movieTracker._trackEvent(…);” This is the part of the code you want to pay attention to:
   

   try {
   var pageTracker = _gat._getTracker(“UA-xxxxxx-x”);
   pageTracker._trackPageview();
   } catch(err) {}

3. Where your code at? This ended up being my problem. People recommend that the GA code should be inserted at the footer of the page to not slow down the loading time of a page. The catch is, the JS file/definition of the variable pageTracker may need to be at the top of the page for _trackEvent to do its thing. If you try moving the code with the ga.js file to the head of the page, it should help. Or it helped me anyway.

As much as I hate debugging JS, I think it has really helped me to have knowledge of how it works, so if this is all kinds of confusing, try working through the W3C tutorials.

Misc.

The Grok delivers!! I was tweeting with @TheGrok and I asked if there was any way he could post about web analytics for non-profits. Brendan Regan wrote an awesomely helpful post about it. I will probably comment more about it later, but still, very cool.

And for those who use Twitter, #wa is being deprecated in favor of #measure. See Web Analytics Demystified for an explanation.

I guess I was feeling too web analytically inclined lately because I have been only reading about IA/UX stuff now.  There was an interesting research posting by ClickTale, that studied how users scroll and whether or not they are actually reading things below the fold.  There are 2 parts to the study, but part 2 is where they give substantial conclusions and advice.

Turns out people do scroll. Go figure. Its relative to the position in the page though, and people’s attention span drops dramatically after the top until they get to the footer, which they pay close attention to. Here are the main recommendations:

• The most valuable web page real-estate is located near the page top, between 0 and 800 pixels. Visitor Attention and Page Exposure peak at about the 540 pixel-line.
• If you have a long web page,  add “stop points” such as headers and images to prevent your visitors from quickly scrolling down the page. It will prevent their attention from waning towards the end of the page.
• The footer of your page is important! Users do pay quite a bit of attention to that area of your page.

This is interesting to me because the primary argument of a person sitting in on a redesign meeting at the Burke was that no one will ever look below the fold, so they asked me to make a different mock-up reflecting that.  While I think the most important/engaging content should remain above the fold, it turned a content-rich page into something that does not have much content because of the design restrictions imposed. You can see an earlier mock-up in the portfolio section, but I was undecided on whether this was a better alternative or not. I would post the new mock-up but I don’t know if that’s allowed yet.

Uber complicated graph from ClickTale

And if anyone is in the market for an amazing smelling candle, I bought the Aloha Orchid one from Anthropologie. Very nice. And all the millions of people who decided to make identical Excel macro tutorials need to stop.

I was minding my own business, looking at some data in Google Analytics, and then I realized that one data set pretty much was a perfect example of the long tail. A) I was excited that I even recognized it when looking at a table, and b) I didn’t know what it really meant, but I did think it would be important to investigate.

General definition: increasingly shifting away from a focus on a relatively small number of “hits” (mainstream products and markets) at the head of the demand curve and toward a huge number of niches in the tail.

SEO definition: lots of key phrases individually account for little traffic by themselves but collectively all those key phrases often could account for a huge amount of traffic.

long tail example

As usual, I first turned to the Occam’s Razor blog for answers, and it seems like the long tail can mean different things in different contexts for analytics.  When you are looking at keywords, and there a lot of keywords with small usage that add up to a big percentage, you would be interested in learning more about the tail. The tail keywords tend to be subject based versus branding keywords. So ours might be “Northwest Indians” versus “Burke Museum.”  These keywords could be better utilized in the content of the site if this is what people are searching on.

On the other hand, I was looking at the top pages for a subsite, which Avinash compares to a “cliff” at the edge of the long tail.  You should be trying to move the cliff further along the long tail ideally.  In that case, most people are looking at the head pages only, and very few reach the long tail pages.

Content Long Tail

The overall gist of the long tail examples was that keyword analysis should focus on the long tail, and with content you can initially focus on the head of the tail.  I guess what I’m having trouble with is how you apply this to other metrics, like average time spent on page, or bounce rate.  Are keywords the primary instance in which you want to initially focus on the long tail to get more insight into user behavior?  How do I decide which pages in the long tail to focus on for top content? If I see another long tail, should I just pretend like I didn’t see it so I don’t get confused again?

Last day of work at the Burke Museum. I am sad I won’t get to finish some projects or keep working on things, but hopefully they get a lot of benefits from what I did while I was here.   Mainly I’m excited to start at ZAAZ after I relax for a few days.

In my last few days at the Burke Museum, I’ve been tasked with doing anything web analytics possible until I leave.  There’s so much more I wanted to accomplish in terms of figuring out goals for the site and assessing who our users are, but hopefully web analytics won’t get tossed aside forever after I’m gone.

We’ve already done overall assessments and reports on site metrics, so I’ve been trying to figure out actions they can take that will be easy to code and maintain.  Dealing with user engagement is the biggest issue so far, especially because measuring it is difficult and open to interpretation. If they at least attempt to address engaging and interacting with our users, I will die happy.

Average Time Graph

I’ve also focused on looking at things that we redesigned recently to track their performance since they went live.  I’m totally a nerd and I got really into how we could track conversions for the Education sub-site, and how we could better market them for next year.  Common conversions for an e-commerce site might be how many people get to a confirmation page after buying something, but for an informative site like the Burke Museum, its more difficult to decide what counts as a user that basically “did what we want them to do.”  Also, the charts I put in the reports are very pretty.

Mainly I’ve just been producing reports that I’m not sure anyone has the time or interest in reading, but what else is there to do when I’m going to be gone soon? I should think of something more creative. They might not read a report, but they will always remember a dirty limerick on web analytics.

…and in other news Michael Jackson died?? Sad times.

I think its only appropriate that in honor of my father, an accountant, that I spent the day BBQ-ing and playing with Excel.  Apparently in web analytics, reporting tools gather the data, but you manipulate or “do cool stuff” with it in Excel.  So I now know way more than I ever thought I needed to about PivotTables and such (which will come in handy at ZAAZ). Excel 2007 does seem to make it pretty easy to do all of the basic things like calculating sums, but I have a sneaking feeling that I will be doing more complicated things than this.

I should probably be focusing equally on data visualization (i.e. charts and graphs), because that’s how I’ll be getting my point across to clients.  Avinash Kaushik says that the best visualizations come from:

“1) a deep understanding of the goal / objectives 2) from thinking beyond what standard trend lines or stacked bar graphs can provide.”

I tried to emphasize my points in my last 2 web analytics reports for the Burke through varied and clear charts and graphs, but I haven’t gotten feedback yet on those.  I think the challenge is in making the numbers tangible without inducing death by bar graph. I tend to understand things better through words, so it requires creative thinking for me to not rely on rhetoric.

First day of summer in Seattle.