Web Analytics - another excuse to procrastinate

Unless it's your job or you are making money doing it, I find a lot of people with blogs and web sites don't often check their web analytics to see how their site(s) are doing.

It's like balancing your checkbook - something you should do regularly, but always put off until the last possible moment. Even the word is foreboding - Analytics.

Well, I'm here to ease your pain. Call your analytics a guest book and you always want to check the guess book to see who has visited or left a message (converted to a sale or download).

Web Analytics is like a computer program - it has all the bells and whistles, but you only need to use the bells or just the whistles. Take a standard word processing program. There are several ways to make a word bold or change a typeface, but you only use one the ways.

It's the same with analytics. You don't need all the information it tracks, only the most important pieces of data related to your site and your promotions.

Let's look at a typical scenario. You send out a monthly newsletter about new products you have just introduced. You know that most of your email list, say 80%, consists of people who live in your city. You have also placed links in your email to only the new products featured.

A few days later a few sales come in and then there is a large spike in sales. Your email marketing firm tells you that 12% of your emails converted meaning that 12% of your email recipients at least opened your email. The marketing firm also tells you a majority of your emails were forwarded to others. But, you don't know where your customers are coming from for future promotions.

You see from your sales records that the spike in sales is coming from Florida and most sales are for a new wooly scarf from a new designer you just took on in your clothing line.

Now you are scratching your head and wondering, "Who would buy a wooly scarf in Florida?"

A quick look at your web analytics and you see that many of your visitors and your conversions are concentrated in pockets in the larger metropolitan areas. And the majority of conversions (sales) were made late at night.

You are ready to call that psychic you always passed on the highway each morning to work to find out who this new customer segment is and their demographics.

But you don't have to call her, just look at your analytics again. Another piece of interesting data shown is that the majority of visitors to your site who purchased a wooly scarf did it from a mobile device. Now you know your new customer segment.

Playing a web analytics detective, you put the clues together:

• A good number of your emails were forwarded to others
• The majority of transactions occurred late at night
• Sales were concentrated around larger metropolitan areas
• Online sales were made mostly with mobile devices

Your new customer segment is mainly composed of female teenagers and young people.
How do you know? Elementary my dear Watson.

When a few teens saw your new scarf, they were excited and wanted to share with all their friends - the email forwarding then ensued.

The majority of sales occurred late at night - teens and young people are busy texting each other at night and probably texted your website address to each other. Ordering was done at night.

Sales were concentrated around metro areas - most teens live close to each other by the location of their high school so you see the higher sales in the more populated areas.

Most of the sales were made using mobile devices - younger people are more comfortable with using a mobile device and especially using it to make purchases. You can assume the majority of sales came from teens and young people.

To test your conclusions, you send another email marketing newsletter with a new scarf and other items from your new designer and track the online sales with your web analytics. Your sales take off.

So you can see it pays to check your web analytics...at least once in awhile. Don't put it off.

The Acid Test

Ok. Here's the acid test. You managed to create a Google Analytics account, generate the javascript code for your site, and place the code at the bottom of your site. So what's next?

Sign into your account and go to your dashboard which will look like this. I'm tracking several sites so yours may have only one listing.


If you look under the status column you should have a green check mark - that means you successfully installed the Google code and Google is receiving data.

IF you have a yellow triangle you're in trouble, but don't panic. Give the site a few hours and come back later to check it. Sometimes it takes awhile for Google to retrieve the data, figure out it's yours, and place it on your dashboard.

IF you still have the triangle after a few hours, you will have to go back and install the code again. Maybe, you placed it in the wrong area. In any event, Google has a lot of help information so take a look if you get stuck.

Ok, so let's assume you have a nice fat green check mark. Now comes the fun part. Really, I mean that.

Click on the "View Report" link and almost like magic, Google has created an analytics dashboard for you complete with a month's worth of data from the day you open the report. Of course, if you placed your code a few days ago you will see very little data. You can change the date range or make it just one day by clicking on the date.



There's a lot you can do here and you'll have to play with it. I'm just going to highlight what you first see for now because I know your brain is starting to hurt again.

The first element is a nice line graph showing the number of visits by day. This shows you which days visitors decided to come to your site and how many.

The next section is Site Usage. This shows visits, pageviews, pages per visit, bounce rate, average time on site, and the percentage of new visits or visitors. Most of this is self explanatory except maybe bounce rate. This shows the percentage of visitors who stayed less than 10 seconds. That means they found your site, glanced through it and left - they were not interested. A higher bounce rate will tell you something on the site is driving your visitors away. Ideally, you want a low bounce rate.

The average time on site tells you just that - how long the average visitor stayed. If the average is over a minute or so you can assume they liked what they saw and read something there.

The last metric tells you the percentage of new or first time visitors. Obviously, you want this to be high because it will indicate that whatever you are doing on other sites, it is driving traffic to your site. But, a high visits number and a low percentage of first time visitors will show the majority of your visitors are loyal and make repeat visits.

Image via CrunchBase

Ok, are you passed out yet? Just a few more metrics to cover and you're off the hook.

Visitors Overview on the left is just that, but the Map Overlay is really cool. Click on any country in dark green and it will show where your visitors are from by country and by state or region. I find most of my visitors are from my home state and that makes perfect sense since most of my friends and relatives are here.

The last two sections are highly valuable in determining what's working and what is not on your site. The Traffic Sources Overview will show you cool metrics like the percentage of direct traffic (visitors type in your site address), the referring sites (where you have links on other sites to your site), the search engines used to find your site and the keywords visitors used in those search engines. All valuable information that will tell you how effective you are in driving traffic to your site.

The Content Overview tells you which content on your site is getting the most visits or interest. This is great for blogs because it will tell you which post is getting read the most and you can determine what subject matter the majority of your visitors like best. It is also valuable for multiple pages on a web site to determine the most interesting pages.

You can also customize the dashboard to show different reports if you don't like the default offering, and you can easily export any report to a PDF, CSV, XML or TSV file in addition to emailing them.

I think that's enough for now. Go pass out.