Google Analytics (II)
- Advanced Analysis -

Basic analysis is mainly a quantitative analysis; if the number is high, is a good thing, if the number is low, is a bad thing, and so on. An advanced analysis is not all about the numbers but rather about how to make correlations between the available traffic metrics in order to make an inference. That's why we will first define the goal that we want to reach (i.e. 'which keyword performs better?') and then we'll work our way back to see which metrics can help us find out if we reached or not the goal.

Keywords Analysis

Ok, so we know which keywords people use in their organic or paid searches. But the question is: which keyword is the best performer? By comparing different traffic metrics for each keyword we can draw conclusions on the efficiency of the keywords and their landing pages. Let us use a real example to better explain he concept. Let's say we are conducting a PPC campaign for the keyword "web host review". By accessing Traffic Sources - Keywords sub-section and selecting [Show: paid] we can see how the keyword is doing.

Analysis:

• The number of visits is a subjective factor as it depends heavily on the ROI of the keyword therefore we ignore it;
• The bounce rate of 37.50% is high (even if it's better than the site's average). This means that you are loosing one third of your visitors on the landing page;
• An almost 55% decrease in the average time spent compared with the site's average. It means that people searching for this keyword are not staying enough on the website. Maybe the keyword used is not that relevant to the topic(s) discussed on the website?
• The number of pages/visit (see below) is again a subjective factor:
  • If the landing page is a sales page, having a low number of pages/visit might not be a bad thing (the visitor bought the product instead of wondering around on the website);
  • If the landing page is an informational page a low number of pages/visit is a bad thing (it means that the landing page is not relevant to the keyword).
• 100% New Visits is not good news. That means nobody wanted to return to the site.

Tip: If you also rank on organic searches for the same key phrase, switch the selection from [Show: paid] to [Show: non-paid]. Compare the two situations and draw a conclusion.

The Content - Overview - Entrance Keywords sub-section takes the analysis even further. You can now see the efficiency of a keyword associated with a specific (landing) page. This is useful when you have multiple keywords ads running that land on the same page or if you have more than one landing page for the same keyword.

Traffic Sources Analysis

By analyzing the three sources of traffic you can identify patters and visitors behaviours. Check the metrics of direct, referring sites and search engine traffic and compare them to the site's average and among themselves.

You can identify which search engine provides the most traffic and which one (if not the same) provides the best quality traffic. You can identify which link partner performs better when it comes to traffic quantity and quality. We will use again a real example to explain the concept. Let's analyze the following two referring web sites.

Referring site #1

It seems that this traffic source is for sure not among of the best ones: low numbers for pages/visit and average time on site, combined with high rates of new visits and bounce rate (people are not interested in what they found and also not interested in coming back - 100% new visits).

Referring site #2

This traffic source seems to be a lot better than the previous one. Not only people stay more on the site but some of them (22%) are actually returning for more. The bounce rate is also at acceptable levels (39%). The quality of traffic coming via traffic source #2 is obvious superior than that of traffic source #1.

Just like with keywords analysis, you can focus your traffic analysis on a single page (go to Content - Overview - Entrance sources). If you have two links on the same external website pointing to different pages on your website, you can find out which link is performing better by looking at their overall traffic metrics.

Content Analysis

Basic analysis showed us the top landing and exit pages. Using the Content Drilldown you were able to make optimization decisions for pages with "low metrics'. But, while knowing how the visitors entered and exited the website can be useful for page optimization decisions, wouldn't be nice (and also useful) to know how the visitors actually navigated on the website? If you could just see the websites through the eyes of your visitors … Well, dream no more. Click on Content - Site Overlay!

Site overlay allows you to navigate your site just as a visitor does, while viewing clicks, conversions, and revenue metrics for each link. In Site Overlay mode you can see the clicks "heat map"; you can identify the most valuable links and areas and reposition some of the links for better click-through-rates.

Goals

Goals are the last pages of a series of pages that a visitor needs to browse in order to reach that defined goal. Sometimes the goal page has no intermediate steps. If the visitors click on a link and arrive at that page, the goal has been reached. Sometimes however, the visitor needs to be "converted", therefore if the conversion process takes more than one page, you will need to define a funnel navigation by listing the sequence of pages that lead to that goal.

Although funnel definition is optional when defining a goal, it is recommended to define one, especially if reaching your goal involves browsing more than one page. The reason is that funnels are useful in tracking how effectively you retain visitors throughout the conversion process (Funnel Visualization). By seeing which pages in the funnel "bleed" visitors, you will be able to take corrective measures.

You can associate a $ value to a goal. Example: by setting the URL of a $10 sale thank you page as the goal, you can set $10 as the goal's value because you know for sure that the visitor could have not reached that page if he had not paid you $10. Use the Goal Value sub-section to track sales revenues.

Benchmarking

One of the latest features of GA is the Visitors - Benchmarking feature. Benchmarking is an optional GA service that shows how your website's statistics compare against other industry verticals. It uses anonymously the data gathered on GA's servers to create some sort of industry benchmarking metrics to which you can compare your own data and see how well or how poor you're doing. But in order to be able to do this, you have to be part of the data collection and enable data sharing in the Analytics Settings menu (see below).

The End

The two Traffic Analysis articles tried to provide you some guidelines about how to compare traffic sources qualitatively and quantitatively, how to identify your best performers when it comes to keywords and landing pages and how to define and analyze goals and funnels. Once again, these are just general guidelines and the actual numbers defining "good and bad" traffic metrics should be set according to website type, traffic size, industry type and so on.