We all know how important metrics like average time on site, number of visitors, visitors by
segments and of course, conversions are. But, there are some other less known metrics that
really can provide some great insight.

Internal Search
The best way to find this information is to set up your analytics to report on what visitors
search on when they get to your site.

The first thing to do is make sure that your search is set up to be trackable. To do this,
just do a search on your Web site. Look at the URL that displays when the results display.

You should see something like (if you have searched for dvd):
www.mydomain.com/searchresults.asp?keyword=dvd

If your results simply show:
www.mydomain.com/searchresults.asp
ask your programmer to switch the form submit method from a POST to a GET (this should be a
quick, simple change).

Once your search results are trackable, then just configure the report.
WebTrends – Paramater Analysis Report
ClickTracks – Internal Search Report
Google Analytics – Site Search

You may be suprised at the big disconnect by what you think visitors call your product or
service, or what you want them to compared to what they actually do.

Robot Traffic
You will need to analyze your log files to get this information. ClickTracks has a good
Robot Report that breaks down your site by individual pages and tells you the:
Number of Visits (Best metric to see how well your site is being spidered overall)
Frequency of Visits (Best metric to see how your SEO results are improving)
Date of Last Visit
Days since Last Visit (Can be confusing because 0 is actually the best result)

Here is a site in dire need of some SEO

Geography
Most analytics packages have some report that tells you where visitors are coming from
broken down by country, state and even further (DMA). Obviously, you must take this

information with a grain of salt due to the limitations of how this information is gathered
(by IP address which can resolve back to an ISP). But, as local search gains more
popularity, it is worthing spending a little time to drill down to a more local level.
Information gleaned from these reports can be used to improve your local pay-per-click
campaigns.

For more information on local search, if you are in the Philadelphia area, you may want
to sign up for Simon’s Local Search session at SearchCampPhilly

To celebrate the upcoming Webinar Wednesday focusing on Analytics, I’m proposing a little contest.

Over the years, not only have I compiled many, many, many monthly analytics reports, I’ve also worked with lots of different marketers, reviewing their monthly reports.

These reports can be very ‘canned’ or highly customized. The reports can serve one purpose (just get it done so it can be checked off the to-do list) or multiple purposes (gut-check how email, PPC and other online marketing campaigns are doing, as well as understanding usability issues.)

What I am proposing, is that you submit your best monthly report to me at jbrazelle@serengeticom.com before the August 27th Webinar on analytics. I will chose the best report and the winner can choose from the following prizes:

Option 1 – Free admission to the Analytics Wednesday Webinar (kidding, everyone gets that).
Option 2 – Free one-hour consulting about analytics configuration, what is important to measure, or you decide.
Option 3 – A Surprise!

The winner can choose how much of the report to share in the post dedicated to the winner.

So, if you are currently a marketer, an analyst, or even a business owner who puts together a monthly analytic report, please feel free to send it on!!

As more marketers are being held accountable for their budgets, proving ROI on campaigns becomes essential.  In order to track ROI, you must be able to track revenue on your Web site.  This is not nearly as complicated or technically difficult (in most cases) than you may think (or have been told).  

There are three ways to track revenue on your Web site:

1.  Using Analytics
2.  Using Search Engine tools
3.  Using Both – Comparing analytics to search engine reporting

Using Analytics
Regardless of whether you are using a java script based analytics programs or analyzing your Web server logs, tracking revenue is completely possible if you have two things:

1.  A unique ‘thank you’ page that displays only when a purchase is successful 
2.  Access to make modifications to your Web site

Without a unique page that only displays when a purchase is successful, you will always ‘over-count’ revenue.  Often times, shopping carts are built using one form that posts back to itself.  You can tell if this is the case because the URL does not change from one step of the checkout process to the next. 

The problem with this is that you never get an accurate count of successful purchases or cart abandonment since the URL is the same.  If your shopping cart is built this way, it is worth it to ask your developer to either add parameters for each step or implement a unique ‘thank you’ page.

If you don’t have access to make modifications to your Web site, you can still track revenue but it will not be as accurate as you will simply be assigning a dollar value to a goal page.  Here is how it is done in Google Analytics and in ClickTracks

Log File Analytics
To track revenue using log files, you will simply have to pass the order total parameter and the dollar value of the sale into the query string of the URL.  As long as your developer knows .ASP (or .ASPX) or .PHP, this should be a relatively straight-forward, quick change.

Or, you can use what is called a beacon which is an invisible image that is used to capture data that is normally not found in log files.  Here is an explanation for doing this with Google Checkout.

Java Script Analytics
The beauty of java script-based analytics is that you can track anything that you want on your Web site.  You just need to plan for it ahead of time, and have the ability to place custom java script on the ‘thank you’ page.  Each analytics package will have a slightly different method (code) for tracking revenue.  But, it should be well documented in the instruction manual.  (A quick search on ‘tracking revenue in webtrends java script’ resulted in this pdf with instructions on page 15).

Using Search Engine Tools
Both Google and Yahoo generate ‘conversion tracking’ code that you can simply copy and paste onto your ‘thank you’ page.  This article also has good instructions for generating the Google Conversion code.  You just have to scroll down to the section ‘Using AdWords Conversion Tracking.’

Also, Nate had written a good article about the improvements that were made to the conversion tracking earlier this year. 

For Yahoo conversion code, click here.

Using Both – Comparing analytics to search engine reporting
I am always an advocate of this method, using both.  Without a ‘gut-check’ in place, it is way too easy to make decisions with bad data  (which we mention over and over).

Understand first that the data from your analytics will likely never match the search engine data 100%.  In fact, depending on the time frame that you are analyzing and several other factors, the discrepancy may be as much as 15%. 

But, by tying your analytic data in with what the search engines report, you will be able to get a very comprehensive picture of what is working and what is not.  And it will be clear what changes need to be made to make your PPC (and other campaign) efforts as effective as possible.

If you are new to PPC or Analytics, feel free to join us on Wednesdays this month for ‘Webinar Wednesdays.

Most Web analytics packages allow you to set up standard reports that are automatically delivered via email, monthly, or weekly. These reports can be pretty useless lists of meaningless data that users tend to ignore or automatically file without paying any attention to them. Web reporting does not have to be this way. Here are some simple tips to take your Web reporting to the next level.

Use context – Long lists of uninteresting metrics are useless. Find interesting data points and explain why they are relevant. Visual reports show where visitors click and give easy, at-a-glance insight to content and usability issues.

Use appropriate images – Using simple graphs and tables can make presenting data more easy to read and understand. Using 3-D graphs or pie charts can confuse the data as they can be misleading or tough to interpret.

Show cause and effect – It is less interesting to know that traffic spiked last month than knowing what caused the increase. Also a high-average time on site may seem like a good sign. But, you may find out that visitors are frustrated and spending a long time on your site ‘ping-ponging’ between pages as they cannot find what they are looking for by taking a deeper dive into the path information through your Web site.

Compare time over time – Comparing year-to-year or month-to-month provides a good way to see benchmark data and spot seasonal trends.

Segment – Probably the most important way to report on your Web site traffic is to break the visitor groups into segments and analyze top metrics for each segment. This is a great way to find out not only what is ‘working’ for all of your visitors, but what content and paths lead to conversions. By segmenting, you can find ways to improve the user experience for all different visitor groups.

Reading Jakob Nielsen’s great article today, ‘Reduce Bounce Rates: Fight for the Second Click’ , got me thinking about delivering my first few trainings when I was with ClickTracks. The reason for this was this heading…

“Unique Visitors” Must Die

Before joining ClickTracks, I felt that Unique Visitors were a pretty important, if not the most important, stat to pay attention to. But, I soon learned the error of my thinking. First of all that metric, depending which type of reporting you are using (log files vs. java script) tends to be wildly inaccurate. (Based on IP addresses in log files and cookies in java script).

But, accuracy aside, it is an even more misleading stat, as Jakob Nielsen explains -

“Chasing higher unique-visitor counts will undermine your long-term positioning because you’ll design gimmicks rather than build features that bring people back and turn them into devotees and customers.”

The funny thing is that people don’t want to hear this. They want to believe that the Unique Visitor count, regardless of any other attribute (e.g. the fact that they leave the site immediately and never return), is the end-all, be-all metric. Countless times when reviewing the concept, ‘Don’t focus on Unique Visitors’ during the ClickTracks training sessions, I was questioned, even challenged.

If Unique Visitors are not the key metric, then what is? This is a great question and can vary from site to site. But, it all boils down to be able to segment out key groups of visitors.

Conversions – What are the business goals of your Web site? There should be more than one. Purchases are the most obvious conversion activities on an ecommerce site, but also important should be lead capture (getting someone’s email address and the permission to continue the conversation with them). Segmenting out visitors who convert can result in valuable behavior information and the ability to improve your site and increase conversions.

Return Visitors/Loyal Users – How many times do visitors return? How frequently? Following the paths of return visitors can give you great insight on the high impact sections of your site. Determining how often and how frequently they visit can give you great insight on the when you need to be updating your content and broadening your inventory.

Word of Mouth Traffic – Is online or offline conversations driving traffic to your site? It is possible to break out the traffic that likely gets to your site from word of mouth. Once the traffic gets to the site, following their path will give you very interesting insight on what people are talking about.

High Quality Traffic – Visitors who spend a significant time on your site (more than 20 seconds). Once again, breaking out this visitor group and following their paths through your site is the only way to find out if they are sticking around because they are happy and finding what they want. Or, if they are hopelessly lost and ping-ponging back and forth through the same pages on your site.

It is well worth getting ‘unstuck’ from the single-focus mentality and take a much broader look at your Web site traffic, not only to get a better understanding of the important traffic but also to be able to improve the user experience.