Google Analytics

Google AdWords

• To get useful information from your AdWords campaign in your Analytics account, you must link your accounts. This is not done automatically.

Conversions

• A conversion (1-per-click) means only a single conversion is reported for a single visitor / customer; this is true even for multiple conversion goals. On the other hand, conversion (many-per-click) reports the total number of conversions for a visitor, following 30 days of an AdWord click.

• To view a detailed Conversion report, including segmentation by conversion name (very useful if you have several conversions), you must include the Many-per-click conversion column. Then you can use the segmentation by conversion name (or type) by creating a normal segment.

Tracking

• With the new asynchronous system, the GA script is loaded asynchronously. There is a queue of commands that can be set; once the script is loaded, they will be executed. Once the script is loaded, any command entered into the queue is also directly executed.
• If you give the "_trackPageview" command twice (to the same tracker), two calls to the Google backend servers are made. For custom variables obtained after a first "_trackPageview", it may thus be necessary to issue a second one. The second call may be with a virtual URL, if the custom variable is associated to the entire visit (session scope).
• To obtain page load metrics, you must add the following snippet:

_gaq.push(["_trackPageLoadTime"]);

It is currently not tracked by default.

Back Office

Concepts

• It is important to understand the difference between dimensions and metrics. Dimensions allow you to select only part of the total data available (for instance: I want data corresponding to page titles), whereas metrics represent the results of the data (how many pageviews, which conversion rate, etc). A good way to distinguish dimensions from metrics is to remember that metrics will always have number values, whereas dimensions are usually text.

• In custom reports, the Explorer choice allows you the full data visualization (with graphs). The flat table choice only reports data in a plain looking table.

• You can plot almost everything over time by using Motion Chart instead of default Line Chart. You have to choose Time as x-axis, with a spanning period (range) of more than one day. To get a meaningful graph you have to check "Trails" checkbox, then play over the motion chart. It would be actually way better if you could just have the graph plotted out without having to check the play button (which does not add anything useful).

Main Interface Entry Points

• Home:
  • Real-time: very useful real time data.
  • Intelligence: useless, allows you to define automatic alerts.
  • Dashboards: contains your dashboards (which is a collection of shortcuts). Not very useful.

• Visitors: information about visitors (browsers, capabilities, country).
• Traffic sources: very important, can help you understand where your visitors come from.
• Content: which pages are viewed.
• Goals: user-defined goals.
• E-commerce: track transactions and items sold.

• Custom Reporting: contains a list of custom reports, and allows you to create new ones.

• Advanced segments: crucial, allows you to set custom segments.

Misceallenous

• To visualize custom variables values or data in the standard GUI, use Standard Reporting -> Visitors -> Demographics -> Custom variables.
• You can share custom reports! Very useful, especially for Kameleoon Analyst.

Limitations

Goals

• Goals can not be set retroactively; you must set them and only once they are set, data is collected.

Custom variables

• Only 5 custom variables. To overcome this, you can put data in the URL (maybe with virtual page views?).
• 50 variables are available in the Premium version of Analytics.

Sampling

• http://code.google.com/apis/analytics/docs/concepts/gaConceptsSampling.html

Questions

• What's the difference between "Organic" and "Referral" ?