My favorite packet sniffer is HTTPWatch (www.httpwatch.com), but Fiddler (www.fiddler2.com) is a very good, free alternative.

The trouble with Fiddler is that it will watch all of your requests, making it difficult to isolate the Omniture requests that you’re looking for.

Fortunately, you can configure a custom filter in Fiddler, which will exclude non-Omniture calls. This makes is more convenient to use Fiddler to look at server requests sent to Omniture. Creating the filter is a little bit geeky, but very easy to do. Here’s how you do it:

1. Open Fiddler
2. From the Main Menu, select Rules > Customize Rules…

3. Add the following code to the class handlers. This adds the menu item, and turns it on by default.

public static RulesOption("&Omniture Only")
var m_OmnitureOnly: boolean = true;

4. Add the following code to the static function OnBeforeRequest. This is the logic for the filter. It tells Fiddler to only include requests that match the pattern 2o7.net. You can change the filter value by changing the value of var re.

if (m_OmnitureOnly){
  var re = /2o7.net/;
    if (!re.test(oSession.host)){
      oSession["ui-hide"] = “hiding boring domains”; // String value not important
    }
}

5. Fiddler will now show you only Omniture requests, as long as the “Omniture Only” rule is selected in the Rules menu.

Once you have configured Fiddler to show you the Omniture server requests, be sure to check out the Debugger white paper in the SiteCatalyst knowledge base for more information on what the different parameters mean.

Share This

While the page view may be replaced as a means of comparing user engagement for competitive analysis (as it rightly ought to), it in no way changes the importance of the page view as a success metric for your own website measurement.

A good analytics implementation will always measure the things that your site’s visitors do that you consider to be “success events”. Generally, success events are those things that make you money. If you’re selling software, it’s purchases and/or trial downloads. For a media site, it’s the ad impression. Traditionally, the page view has stood in as a substitute for ad impressions as a success metric for media sites because ad impressions are tracked through different systems, and it has historically been difficult to integrate ad server data into web analytics tools. However, that is changing, and web analytics vendors now make it easier to upload data from external sources.

A good analytics implementation also measures the user behaviors that lead up to the success events. Path analysis, in its various forms, is essential to understanding a user’s navigation through your site and the things that lead up to a success event. For media sites, path analysis is even more important as we try to understand what the most popular types of content are, and how they relate with each other to increase site “stickiness” (increased time on site), and “loyalty” (repeat visits). The page view is central to path analysis, and even though our definition of what a page view is may change as a result of rich media in its various forms, you’ll still see the page view as central to your path analysis.

On the other hand, as our ability to track site success more specifically increases, the value of the page view as a key performance indicator may diminish. We’re all trying to measure our site’s return on investment, and as we are more able to track those things that directly make us money (display ad impressions, pre-roll video ads, transfer to affiliate sites, etc.), the page view is no longer as necessary as a proxy for those things. So instead of “Page Views per Visit” as a key metric, we might find ourselves looking at “ad impressions per visit”, or even better - “revenue per visit” as our key measurements.

But remember that the page view is still a critical component of understanding user navigation and as a way of inferring users’ interests. So even if we rely less on the page view as a KPI, it’s still one of the central measurements of user behavior overall.

As rich media confuses the “definition” of a page view, it becomes difficult to provide a standard comparative measurement that applies consistently across different websites. So competitive measurement must begin to rely on those other success measurements, such as time on site, or market share of visitors. I expect that trend to increase as technical innovation continues to confuse the technical definition of the page view, and Nielsen is certainly leading the way in this trend. It will be very interesting to see comScore’s reaction to Nielsen’s announcement.

In sum, I still think that the “page view is dead” idea is over-hyped. Even though I suspect that the page view’s days are numbered for KPIs and competitive analysis, I expect the page view to mature and evolve rather than to go away as an important part of website measurement and optimization.

Share This

http://tech.groups.yahoo.com/group/webanalytics/

Share This

In my opinion, good web analytics involves both the process of analysis and the process of synthesis. According to dictionary.com, analysis can be defined as:

1. the separating of any material or abstract entity into its constituent elements (opposed to synthesis).
2. this process as a method of studying the nature of something or of determining its essential features and their relations: the grammatical analysis of a sentence.

Synthesis, on the other hand, is defined by dictionary.com as:

1. the combining of the constituent elements of separate material or abstract entities into a single or unified entity (opposed to analysis).
2. a complex whole formed by combining.

Even though analysis and synthesis are considered to be opposites, I find that both are essential to what we do as web analysts. Analytics tends to bring together the opposite ends of all kinds of spectrums. Think about it; business users and web programmers, answers leading to questions, quantification and qualification, online and offline. There is virtually no other line of work that offers such a wide variety of business decisions, practices or personalities.

So here’s why I think that both analysis and synthesis are essential to web analytics:

Analysis

• Finding sources of traffic - where did our users come from?
• Splitting a site’s structure up into a taxonomy, or hierarchy - what parts of our site are the most popular?
• Dictionary.com’s definition 5e is “a method of proving a proposition by assuming the result and working backward to something that is known to be true.” - how many times have you been asked to do that?!

Besides, we call it web analytics, of course analysis is involved. Besides, we deal with numbers a lot. But synthesis? Consider:

Synthesis

• Correlating user behavior - what are the most popular pages for our English-speaking users from Canada?
• Consolidating online with offline data - when we run an Internet campaign, do our offline sales increase?
• As I mentioned above, we bring together all kinds of different people, from the geek to the marketing agency hipster (who secretly want to be geeks anyway)
• Creating personas out of various user attributes and navigation behavior
• Combining numerous user properties and navigation elements to decide which advertisement or article to show to a user on the fly (behavioral targeting).
• Putting together various bits of data to create KPIs, then putting all of those KPIs into a single dashboard report for Executives and other decision makers.
• Making sense of seemingly disparate information (analysis) in order to drive a successful business decision or strategy (synthesis).

So hopefully, I’ve won you over. It seems like a lot of thinking around something as simple as a blog title, but I hope to keep blogging for a while, so it was important to come up with something that made sense and communicated something about what this little piece of Internet real estate is all about!

What do you think? Is this name lame like my puns? Insightful? Do I have too much time on my hands, or is it worthwhile to consider these small things every once in a while? Let me know.

Share This