Using Analytics on Your Site

If you're investing time or money in SEO/PPC, you need to know what you're getting in return. The best measure is your bank account and ROI, but with careful analysis of your visitors and their behaviors, you can increase your bottom line, help redirect efforts to the most profitable website segments, and get more out of each landing page. In this article we discuss web analytics and its usage for your website.

Analytics Products

There are many tools on the web for free and others for reasonable prices. Take your pick.

• Google Analytics - Google has a lot of built-in advanced features comparable to expensive analytics tools. The down side is, when you use Google Analytics, you're sharing all your data with Google, and they can do anything with it. For instance, if a certain keyword converts well for you, logically it should convert just as well for other companies, so Google may raise minimum bids for that keyword to make more money.

• Sitemeter

• Omniture

• Efficient Frontier

• Yahoo Web Analytics (free, expected to be out soon at the time of this writing).

• Mint

• Clicky

• Clicktracks

• Microsoft adCenter Analytics

• WebTrends

• Piwik

Once you set up your tool, it's time to analyze your traffic!

Here we discuss basic web analytics measurements universal to all analytics software programs.

• Visits - total number of visits to the website.

• Page views - number of all page views in total.

• Pages Per Visits - average number of pages users sees per visit.

• Bounce Rate - percentage of visitors who come to your website and leave without clicking on anything.

• Average Time On Site - how long people stay on your website

These are the core metrics in all analytics programs. They give you an overall look, but to find real value you have to dig a lot deeper into statistics such as monthly comparisons, trends and segmenting, which can sometimes answer the "why" question.

Traffic Sources

Understanding traffic sources is straightforward. There are search engines, PPC search ads, contextual networks, direct traffic, RSS readers, email, and affiliate/link referrals. Each traffic source can hint at intention, and measuring sources is essential to finding out your strong and weak spots.

For instance, if most of your traffic is from search engines, with a small percentage of direct visitors, chances are you have a weak brand. Focusing on brand building can thus increase direct traffic. Let's take a deeper look at traffic sources:

• Search Engines - Google,Yahoo, Live, MSN, AOL, Ask, Altavista, AlltheWeb, Netscape, etc.

• Referring Sites - These include directories, links from other websites,blogs and banner ads. By exploring referring URLs you can sometimes learn why users clicked on the link. If particular links send you a lot of qualified traffic that converts, you might want to establish a commercial relationship with that site owner.

• Direct Traffic - Those are visitors who directly type in your address or land on your pages through a bookmark. This is a sign you have a strong brand. Direct traffic is free and usually converts well.

Bounce Rate

Bounce rate shows how many visitors land on your website (PPC, search, direct, etc), and then click the back button or close the window. Basically they leave without doing anything. The down side of this measurement is that the bounce rate on a site-wide level does not tell you much. You have to drill into each page, measure its bounce rate and only then make conclusions.

Page Level Bounce Rate

Why is measuring bounce rate on the site wide level wrong? You may change the wrong stuff. The site wide bounce rate is an average sum of the bounce rate from all of your pages. One page can be doing well, while another one is not. How do you know which one is doing well and where you need improvement if you only look at your site wide bounce rate?

By exploring each page one by one and by studying bounce rates, you can spot pages that are doing well in terms of this measurement and ones that need changes.

This is one of my favorite measurements, as it tracks the entire performance of the website. All investments in SEO, PPC, design, copy writing, content, and conversion optimization come down to this statistic. How many visitors do what you want them to do? How many subscribe, buy and become customers?

In this section we'll focus on goal tracking (conversion tracking) with Google Analytics.

"A goal is a website page which a visitor reaches once they have made a purchase or completed another desired action, such as a registration or download." - Google Help

Setting Up Goals in Google

To set up goals in Google Analytics you must meet a few requirements:

• There must be a clear URL for the goal page, such as a thank you page. Basically it's a page that users see after they do what you want them to do. This page should be only available upon completion of your goal. If it can be accessed otherwise (through search results, links, etc), then conversion results will be inflated.

• You must make up a name for your goals. For example, "registration" or "sale."

• You can specify a funnel to the goal page. The funnel represents the page flow before your visitors become customers. In e-commerce, this is the checkout process. In lead generation, this is the application process. Once you specify the funnel, Google will track goal completion. By analyzing goal completion you can learn the exact step in the conversion process that hurts your bottom line.

• Assign a value to your goal. A goal's value helps you estimate the ROI delivered with each conversion.
  
  Google says that a good way to value a goal is to evaluate how often the visitors who reach the goal become customers. If, for example, your sales team can close 10% of people who request to be contacted, and your average transaction is $500, you might assign $50 (i.e. 10% of $500) to your "Contact Me" goal. In contrast, if only 1% of mailing list sign-ups result in a sale, you might only assign $5 to your "email sign-up" goal.
  
  I personally think this is too much information to share with Google, since it can easily estimate how much you make with your website. As it measures goal values on other competing sites in your industry, it can create ROI benchmarks and use that data to price fix their AdWord bids with the excuse of quality scores.

Setting up your goals is easy. Keep in mind that it's essential to redirect users to the goal page once they have completed your action. Without a goal page, Google cannot measure your conversion rate. To set up goals, go to Analytics Setting and click on "edit" in the Actions tab (far right). Click on "edit" next to the goal and enter the required URLs along with the appropriate information.

Segmentation is a huge and complex topic. We'll introduce you to the overall concept and provide further direction where you can learn more, from people who understand it better. Segmentation is powerful and can help you answer the "why" question when it comes to online marketing efforts.

What is segmentation?

Website statistics is a mix of visitor intentions, visitor sources, behaviors and questions. By looking at overall website statistics you cannot find out how many visitors who were looking for "X" found it, and if they didn't, why not. You cannot answer why visitors did not get to "page Y" and what you can do to help them get there. With segmentation, however, you can break down analytics into small chunks of information that can help you market better.

For instance, if you've invested in branding, naturally you want to find out how many people search for your brand, and how many end up on your site and take action. To do this you can create a custom segment, and set a filter for "my brand keyword" (including related). You can go further and only count second time visitors, who spend X amount of time on the site, complete some sort of action or make X number of clicks. By setting advanced filters you can see how your campaign efforts affect visitor flow, engagement and ultimately conversion rate.

Here is a big point to keep in mind: you must know what it is you're trying to find out from segmentation before you segment. At the moment, Google has around 100 segmentation options, so you can waste hours just playing around with those. Know what you're trying to measure.

Here are some Google segmentation options: hours of the day, page depth, visitor type, count of visits, city, language, region, ad group, keyword, search engine, ad slot, referral path, medium, page title, host name, refined keyword, landing page, exit page, affiliate, city, product, product category, browser, connection speed, flash, entrances, bounces, purchases and a LOT more.

To View the Original Article, click here