MEASURING EFFECTS: WHAT, WHEN AND HOW?

Analyzing website statistics to measure which keywords work

What are we measuring? Website data and statistics. How are we measuring? By using website analytics tools. When do we measure? From the first day you start running an SEO/SEM campaign. Why measure? To verify which keywords perform best, what changes caused what effects and what should be copied across other campaigns.

To measure effects of all SEM (and SEO) campaigns, website analytics tools are required. While there are multiple website analytics applications around, from free tools like Google Analytics, to paid applications like Mint or WebTrends, all web analytics packages fall into two categories:

• Server side (gather data from 'log files' located on the server)
• Page tagging (gather data from cookies and directly from the site)

Choosing one application type over another depends solely on the site owner's preferences. Both types have their pros and cons. So, we encourage you to learn about them prior to making your decision.

Server side analytics

An advantage of log file based applications is their ability to harness vast amounts of detailed data. What is more, log file based applications don't require any changes to the site's source code. Log file applications gather and locate data directly on the server. Furthermore, that data is archived and databased on the server in a universal format, making it easily compatible with various web analytics applications.

The con of log file based applications is that they tend to be expensive at start-up, running and maintenance. Because they collect and store data, memory space soon becomes a problem driving up costs. Data needs to be stored for years in order to be useful in seasonality and forecast analysis. However, with each year storage costs grow exponentially. Also, in order to use log files and correctly analyze data, one needs to constantly update web analytics packages with new enhancements.

A hidden, and often undermined disadvantage of log file analysis is the amount of data it produces. For inexperienced users, the boat loads of statistics gathered by these applications can easily confuse and cause information overload.

While information overload may sound silly, it is a serious matter. Consider that your application gives you the ability to measure: every click, of every day, of every second, on every page, of every visitor, split by: sex, age, location, screen size, keywords used, links clicked, time on page etc., all across multiple domains, it's not hard for some users to get entangled and lost. Most often, people simply loose track of what they are trying to measure, what is the most important metric to test, what should be analyzed and what should be compared to... what.

That is why many companies, which utilize complex website analytics applications also hire datamining specialists, or outsource this task to companies that process and analyze their findings, grouping all reports in an easy to swallow format.

Page tagging

An advantage of page tagging based applications is that they don't require purchasing and implementing expensive and complicated software. All that's needed is an inclusion of some JavaScript code to the pages that you are interested in measuring and analyzing.

With page tagging applications, you can easily analyze relevant data like:

• visitors screen size
• most popular priced products bought by visitors
• the extent to which contact forms were filled out
• whether visitors dragged their mouse over a link, image or a portion of the site but didn't click, and so on

Possibly, the most important perk of page tagging is the ability to measure website statistics by companies that don't have direct access to their site's servers.

A disadvantage of page tagging lies in less data that it gathers, especially deep data in comparison to log files. Page tagging also requires tempering with the site's code in order to add scripts.

Which application to choose?

Page tagging and log file based applications are priced at different levels and can be either expensive, cheap or even free. There are many web analytics tools online. All you need to do is find one priced to fit your budget and make sure it suits your needs.

For many companies and private site owners, the most cost-effective application so far is Website Analytics offered by Google. Website Analytics is a page tagging application, and is completely free.

For amateurs, beginners and even some small to mid-sized companies, Website Analytics has enough capabilities to suit their needs. Regrettably, because it doesn't track IP addresses, it lacks some segments and doesn't measure some deep data. Its goal analysis is also limited in scope. Nonetheless, we recommend using Website Analytics if you have no other application and are interested in measuring your site's performance over time.

Another application worth mentioning is Mint. Mint shows real-time data and website statistics along with collecting IP address information (thanks to which you can trace each of your visitor's visits across your whole site). An important part of Mint is its ability to showcase keywords that are currently used by visitors to find and access your site.

You have a website analytics tool? Well, let's analyze

To measure performance of your site or your campaigns, you must first decide on their goals. Of course, the objective of most campaigns and sites is to raise sales, but because of market specifics, some companies measure performance as the amount of contact forms filled out, or downloaded documents, etc.

It is imperative to clearly define what your site's and campaign's goals are, in order to accurately analyze what impact different variables have on them.

Once you choose your objectives, start gathering data such as:

• the amount of visitors who did not purchase (checking at what point, or what page they abandoned the sales tunnel)
• what keywords attracted most buyers
• pages, which visitors who abandoned the sales tunnel landed on
• what were the most popular links clicked, on what pages, by visitors that filled out contact forms
• what pages convert most traffic (check what content is read by visitors that end up purchasing)

You can measure many variables across many segments. The important thing is to measure paths chosen by visitors, especially visitors who buy. In so doing you will find pages that are most clicked on - which tend to be your best converting pages.

By analyzing data you can determine what changes and modifications are needed to compress the sales tunnel. By compressing the sales tunnel, you can ensure that visitors reach appropriate services/products or contact forms, taking action that you want them to take.

When analyzing pages or campaigns, pay attention to your keywords list and check which keyword, phrase or ad performs best.

Check:

• which ad converts best
• which keywords attract most traffic
• which phrases are expensive, yet ineffective
• which phrases are competitive and produce maximum payback at minimum cost

With proper analysis and action, you'll eliminate many SEM traps and obstacles from your visitors' path. Thanks to proper analysis you may be able to optimize the site to its maximum capacity and performance.

Unsure where to start?

Contact us for a consultation. Have SEO Magma answer any question you might have about SEM and about building successful AdWords campaigns.

If the ideas presented above are unfamiliar to you, then we invite you to read the Purpose of website positioning: how to use SEO and SEM where we explain SEO, SEM and the differences between them in greater detail.

^ back to top