It’s not all about Pay per Click

Thursday, April 23, 2009 | 5:28 PM

Labels:

Guest post by Barbara Pezzi, Director Webmarketing and Ecommerce, Swissotel

In many companies web analytics is mainly used to track clickstream traffic, or at best, to track pay per click campaigns. But there is so much more to it. Whilst tracking the return on investment of web marketing campaigns is definitely a must, unless we understand how our site visitors interact with our pages and adapt our content accordingly, even the best marketing campaign will ultimately just bring additional traffic, which will fail to convert due to poor content.

In order to truly understand the site experience from our customers’ perspective, it is essential to segment them into relevant groups and give the figures some context. Every website is a mix of visitors, sources, and behaviour. Looking at the data in aggregate means nothing. The recent introduction of advanced segmentation in Google Analytics makes it really easy to “group” our visitors into relevant segments and understand what they were looking for and what they enjoyed.

We are no longer limited to tracking paid vs organic traffic, but can now create our own segments, relevant to our business needs, in just a few clicks:
  • visitors who stayed less than 10 seconds
  • visitors who booked more than once
  • visitors from a specific Google AdWords campaign
  • visitors from a specific country or city
Let’s take, for example, the Swissotel Merchant Court in Singapore. 50% of our website visitors are from Singapore. This is an important segment. We need to be able assess the unique characteristics of user behaviour from Singapore if we are going to be able to optimise our content and marketing for this important audience. With just a couple of clicks the “Visitors from Singapore” segment is created.


I can now see what pages are popular with our visitors in Singapore (in the example below, 90.53% of the traffic to the hotel restaurant page is from Singapore), what keywords are used to find the site, what promotion is working, search engine used, and so on.



Based on the data I can now add more relevant content, choose the right search engine and keywords when I run specific PPC campaigns for Singapore, and create additional packages which will appeal to the market, even for off-line sales. And when I run my next campaign targeted at Singapore, I will create a new segment “AdWords Singapore Visitors”, compare it to the “Singapore Visitors” segment, and start the exercise again. Web analytics is a continuous cycle, and with advanced segments the possibilities are endless.

I am always on a quest for real actionable data. The actions we take based on this data more often than not result in a more enjoyable site experience for our customers, and in return, higher conversions and repeat business for us. Advanced segmentation is a powerful and very easy to use feature, but extremely addictive…

5 comments:

Kati said...

Hi Barbara,

really interesting - well done

Cheers Kati

Dr. Wisdom said...

Hi,

I need a report on the click stream through different referrers. Now, each time we have different referrers so I can't simply filter and create different profiles for each of them.. IS there a report on GA ?

Vinny said...

Hi Dr. Wisdom,

The good thing about Advanced Segments is that it minimises the need for you to create dedicated profiles and filters for each segment that you are interested in.

To discover all channels driving traffic to your site, take a look at the Traffic Sources > All Traffic Sources report. You can then create an Advanced Segment for each source or collection of sources that you are interested in.

Vinny

carsten said...

Hi Barbara,

nice article,

you say that you can set up a segemnt wich showes:

"visitors who stayed less than 10 seconds"

Can you explain me how to do that?

Vinny said...

Hi Carston,

These steps will help you create an advanced segment for "visitors who stayed less than 10 seconds":

1. Create a new advanced segment
2. "Time on Site" "Less than" "00:00:10"

Note that time on site for single-page visits are always calculated as zero minutes.