Analysis posts – Analytics Talk

Tracking Internal Campaigns with Google Analytics

March 30, 2010 By Justin Cutroni 19 Comments

Internal campaigns are marketing efforts that are run on your site and promote your products and services. Here’s an example from the Boton Red Sox site. They’re using ads on the homepage to promote ticket sales.

Companies should track how people react to these campaigns and which ones are most successful. But what’s the best way to do this with Google Analytics?

Some people use the standard campaign tracking to track internal campaigns. THIS IS INCORRECT AND SHOULD NEVER BE DONE. Using the standard campaign tracking for internal campaigns will cause problems with your source data. So don’t do it!

There are a few correct ways to track internal campaigns. You could use Event Tracking, Custom Variables or Virtual Pageviews. But I like to use GA’s internal campaign tracking tool.

What? You’ve never seen or used the GA’s internal campaign tracker? It’s in the profile settings and it’s called Site Search tracking! Did I fool you ;)

Site Search can easily be configured to track internal campaigns. Let’s walk through the steps to set it up and then the data and analysis.

Step 1: Create a New Profile

Because we’re using Site Search for an unintended purpose it’s best to configure these settings on a new profile. It’s not possible to use Site Search for both tracking internal campaigns and internal site search within the same profile. You need to have a separate profile to track internal campaigns.

Step 2: Tag your Internal Campaigns

Once you’ve created your new profile it’s time to tag your internal campaigns. Internal campaigns need to be tagged in a similar manner to external campaigns: you need to add query string parameterrs to your internal ad.

However, unlike external campaigns you do not use the standard link tagging parameters (utm_campaign, utm_medium, etc.). You get to make up your own parameters!

You can use one or two parameters for internal campaign tracking and you can name then anything you want. The reason you can use one or two parameters is that GA’s site search configuration uses two parameters, one for the search phrase and one for the search category.

Whatever you choose, make sure the parameters are not used for anything else.

TIP: Check your Top Content report for a complete list of your site’s query string parameters. Verify that the parameters you create are NOT in this list.

For the sake of this post I’ll use the parameter icn (shor for internal campaign name). This parameter will holds the name of the internal campaign. I’m going to use the following format for the value of the campaign name parameter

icn=[internal-campaign-name]

I mentioned that you can use two paramters. You don’t need to use two, but GA’s site search can be confiugured to track the internal site search phrase and a site search category. We’ll use the category paramter to track the internal campaign name.

I’m going to name the second paraeter ici (short for internal campaign info). Again make sure the parameter you’re using does not already exist. This second parameter let’s me collect details about the ad the visitor clicked on and the location of the ad.

Here’s a basic format:

ici=[ad-creative]_[location-on-the-page]

You can see that I’m stuffing a lot of information into the parameter. You can put whatever you want and GA will gladly suck it in. By adding more information we’ll get a granluar view of how the internal campaigns perform and which locations and variations lead to tbe most conversions.

If you don’t have different types of internal ads, or just don’t care about this level of detail, then you can ignore the add internal campaign info parameter. It blank, it’s up to you!

Now you need to define the values for all the ads. Thic can get messy if you’re running a lot of internal campaign. But you can do it, just be organized! Use a spreadsheet to keep track of all the values you use.

Once you’ve got al your parameters it’s time to tag your links. The exact process depends on your site. You may need to change static links, like this:

< a href=”/internal-page.php?icn=2010-spring-sale&ici=stubs_home-roller >

Or if you have complicate flash ads you may need to get inside the Flash code. It depends on your site.

The bottom line is when somone clicks on an internal ad you want to see your internal campaign parameter on the next page. If you don’t see the parameter in the URL then you did something wrong.

You can use the sample spread sheet below to track the different parameters you use for your internal campaigns. The spread sheet also has a formula in column D to automatically add the parameters to your URLs.

NOTE: There is an iFrame in this post. If you can not see it, you can view the original post here or view the Google Spreadsheet here.

Once youe’ve got the parameters added to your links it’s itme to configure the Site Search settings.

Step 3: Configure Site Search Settings

Remeber, we’re configuring these settings on a new profile so we don’t break the site search in our main reporting profile.

Site search has three settings. First, turn site search on.

Next, tell GA the name of the paramter that holds the site search phrase (in this case it’s out internal campaign name) by adding the parameter to the ‘Query Parameter’ filed.

Next, choose Strip Query String Parameters. This setting will remove the parameter from the URL after GA processes the data. This is a good idea because it reduces duplicate pages in your top content reports.

TIP: You probably want to exclude your internal campaign name parameter, and internal campaign information parameter, from your other profiles. It can really mess up your pageview data.

If you’re using an internal campaign information parameter configure the Site Search Category settings the same way. Just make sure you use your internal campaign info parameter in the ‘Category Parameter’ setting.

Here’s how the settings look using the parameters from my example:

That’s it! Let’s look at the data.

The Reports

Let’s start by answering a simple question: do people who respond to internal camapigns convert more or less than those that do not respond to internal camapigns? To answer this question use the Content > Site Search > Usage report. Here we can see that there were only eight visits that clicked an internal campaign. Sad! But it’s just test data.

Now let’s drill deeper ad identify which inernal camapigns are most effective. Use the Content > Site Search > Search Terms report. Rather than search phrases this report contains the names of all internal campaigns. Again, what was the response to the campaign? Was it worth the effort? Don’t forget to check the Goals tab and the Ecommerce tabs (if applicable) to measure outcomes!

But let’s drill deeper to understand which ads within those campaigns are working. Click on a campaign name and choose Category from the Analyze drop down.

Now we’re looking at all of the information that we put into the ici query string parameter for this particular campaign name. If we had multiple internal ads we’d be able to differentiate ad placements and creative variations.

Don’t forget to use the Goals and Ecommerce tabs to measure outcomes! This is what most people want to know: did internal campaigns, and specifically which internal campaigns, generated revenue and conversions?

But we can do more. Now change to the Content > Site Search > Start Pages report. Now you can see which page people were on when they click on an internal ad. Again, more insight into where visitors responded to an internal campaign.

And for all those marketing folks that are so concerned with internal campaigns, how about creating a nice custom report and automating the delivery or, better yet, use the Custom Report Sharing feature to share this report with others. People will love this because you can change the wording so it does not say Site Search it says Internal Campaigns Report.

But wait, there’s more! What about using a secondary dimension to view the external marketing campaigns (or sources, or mediums) that drive visitor to react to internal campaigns. Perhaps the extrnal creative has some influence over how visitors react to the internal campaign creative. The data isn’t so hot in the image below, but you get the idea.

And finally, the ultimate in analysis, internal campaign attribution. We can use the Search Term Refinement feature if visitors click on multiple internal campaigns. Google Analytics will track all subsequent site searches, but in our case follow up site searches are actually additional internal campaigns that the visitor responded to. Honestly, I have never found any insights from this type of analysis, but you can do it if you want!

Ok, I’ve officially entered nerdville.

I think you get the idea. By adding all this data you can do many different kinds of segmentation and analysis. More than enough to understand the behavior of your site visitors and how your internal campaigns perform.

Last but not least, I’ll mention that you can track internal campaigns using events and custom variables. But both of those solutions require coding. And that requires working with IT. Using Site Search, in most cases, will not require any code changes to your site.

Share this post:

Filed Under: Analysis, Campaign Tracking, Tracking Tagged With: Analysis, Campaign Tracking, site search

Google Tackles Campaign Attribution with AdWords Search Funnels

March 24, 2010 By Justin Cutroni 4 Comments

There’s been a lot of debate in the analytics community about campaign attribution and how to assign value to the various marketing touch-points that lead to conversions. If you’re new Campaign Attribution you should check out the book Web Analytics 2.0, it has a good, functional overview of the attribution challenge.

Throughout the discussion it has become clear that the classic first click and last click attribution models that many web analytics tools use are flawed. The problem is no one has come forward with a better solution to the attribution issue… until now.

Google has taken a very low-risk move by tackling campaign attribution for AdWords only. The new AdWords Search Funnel reports help marketers understand which cpc ads people see and click on prior to converting.

If you’re looking for details about the reports and how to use them check out the video below from Google. The new Search Funnel reports have not been rolled out yet so no one has had a chance to play with them. Hence no real description here :)

We’ve long known that people see a lot of different cpc ads during a sales cycle. Avinash Kaushik calls these keywords “upper funnel” keywords. They are used by people that are early in the buying cycle. While many of these keywords don’t always lead to a conversion they help educate a potential customer and move then closer to purchasing a product or service.

Even though they do not directly generate revenue there is some value in bidding on upper funnel keywords.

Up until now we haven’t had many ways to help us understand the true value of upper funnel keywords. Sure, we can use time on site or pageviews per visit to measure “engagement”, but that was a bit of a hack. We can also create all sorts of custom JavaScript to store the first click and last click in a Custom Variable. But again, these are just hacks.

The Search Funnel reports are a well thought out way to understand how people interact with AdWords ads prior to conversion and thus help us understand the ROI of our AdWords spend. The reorts provide insight into which keywords

I think this is a good first step by Google. They took reliable set of data that was just sitting around a data center and created some reports that will help marketers understand the real value of different types of keywords. This is all very low risk for Google with very high potential (read: more AdWords revenue).

But these new reports are also a good test of how users, and the overall analytics market, will respond to Google’s version campaign attribution reporting. Real attribution models are very complicated to create. They involve a lot of data about different types of campaigns (banners, cpc, email, etc.).

[Side note: Why is it that we haven't seen any DoubleClick data in Google Analytics yet? Pulling that data into GA will be critical for real attribution measurement.]

In addition to the data complexities, every business will have their own way to weight certain marketing activities in an attribution equation. For example, some companies may value email more than paid search. This business logic will be difficult to implement. Not impossible, but difficult.

At the end of the day the new AdWords Funnel reports are exciting. But I’m excited to see how Google takes information about how these reports are used and tackles the bigger challenge of true campaign attribution!

Share this post:

Filed Under: Analysis, Campaign Tracking, Web Analytics Tagged With: adwords, Analysis, attribution

New Google Analytics Goals

October 20, 2009 By Justin Cutroni 44 Comments

We all know that it’s critical to measure conversions, or goals, for our website. But for a long time Google Analytics limited the number of conversions, and types of conversions, you could track with Google Analytics. All that changes today (October 20, 2009).

You can now create up to 20 goals per profile in Google Analytics. I can literally hear the applause at eMetrics :)

In addition to expanding the number of goals Google has expanded the types of goals to include ‘threshold’ goals for pageviews per visit and time on site.

I think we all know the importance of tracking goals, so I’m not going to get too deep into why you should use goals. If you’re not using goals you should start NOW!

Let’s talk about this new feature.

Goal Sets

Goals are now organized into four sets. Each set of goals can contain up to five different goals.

Google Analytics Goal Sets

Sets have been introduced as a way to accommodate all the new data in GA. In the report tabs, rather than one goal tab there can be up to four goal tabs in a GA reports.

New Google Analytics Goal sets in a report

When creating a goal you can place it in any set as long as there is room. Once you place a goal in a set it’s best to NOT MOVE IT. Google Analytics sees this as a NEW goal and does not move the previously captured conversions to the new goal.

TIP: I like to organize goals by business function i.e. put goals that are related together. For example, if you’re a content site, you might create goals for spending a certain amount of time on site (1 minute, 2 minutes, etc.). I would group these goals in a set all related to time.

Goal Types

In the old days a goal was a pageview that represented the completion of some high value process, like a thank you page. Now goals can be based on actions that have nothing to do with viewing a page. Conversions can be based on how much time a visitor spends on the site or how many pages the visitor views.

Time Based Goals

Time based conversions are triggered after a visitor has spent a certain amount of time on the site. To configure a time based goal enter the hours, minutes and seconds that a visitor must spend on the site before a conversion is counted. Once the visitor reaches that amount of time on the site then a conversion is triggered.

Creating time based goals in Google Analytics.

What’s interesting here is that you can create a time based goal if a visit does NOT reach a certain amount of time. If you choose ‘Less Than’ Google Analytics will trigger a goal if a visit does NOT reach a certain length.

Less Than Goals in Google Analytics

Why on earth would you measure this? I like to think of ‘Less Than’ goals as ‘Failure’ metrics. We often define success metrics, like Conversion Rate, but rarely define metrics to measure our failures!

Using failure based metrics really packs a punch when you’re talking to co workers or clients. For example, when you configure a failure goal you can easily measure and say, “Did you know that 97% of our traffic does not spend at least 2 minutes on our site? We suck!”

Abandonment rate is another well know failure metrics.

Time based goals can also be very useful if you’re trying to MINIMIZE the amount of time people spend on your site. For example, if you have a support section on your site you may want to understand what percentage of traffic spends a certain amount of time on your site. Long term you can try to reduce the number of visits that are too long.

How about setting up a goal set for various time intervals and then try to move visitors from one “goal” bucket to the next. 10 minutes, to 7 minutes, to 5 mintues… You guys are bright, you get the idea :)

Remember, time based goals can be affected by creating virtual pageviews and events. Both of these activities send data to Google Analytics and can change how visit length is calculated.

Pageview Based Goals

Another new goal type is pageviews per visit. Like time on site goals this this type of conversion is triggered when a visit exceeds a certain number of pages. I can literally hear all the advertisers clapping out there!

Pageviews goals are set up in the same manner as time based conversions. Just specify a condition (greater than or less than) and the number of pageviews in a visit.

Pageviews per Visit Goals in Google Analytics

Like time goals, pageview goals can also be affected by virtual pageviews. If you’re creating a lot of data using _trackPageview() you need to understand that this can change your overall goal calculation.

URL Destination Goals

The old standby! ‘Traditional’ goals are now called URL Destination Goals. You can still use a regular expression, head match or exact match to identify a page that represents a goal. This functionality has not changed (you can learn more about goals in this old post.)

URL Destination Goal in Google Analytics

Now that we have 20 goals we can easily measure all of those micro conversions (RSS subscription, email signup, reaching product page, downloading white paper… etc, etc, etc).

And yes, you can still use a virtual pageview as a URL Destination goal.

Funnels

Google did spend some time tweaking the interface. The old interface always showed 10 steps in the funnel. Now you can choose the number of fields the funnel form displays. You’re still limited to 10 steps in total. This isn’t such a big deal.

New Funnels interface in Google Analytics

But think about the bigger picture. Do we really need funnels if we have so many goals? With 20 goals we can use a goal to represent each stage in a process, rather than a funnel step? So do we still need funnels?

Yes. Funnels provide a nice visualization of critical processes, so I think they are still relevant. Plus, you need to configure a funnel if you want to measure Abandonment rate, a very nice failure metric that can make people squirm :)

Odds and Ends

A few random thoughts re: new goals:

If you’ve been creating lots of profiles for goals you may want to consider consolidating all goals to a single profile. The benefit is you can have all your conversions in one interface. No more messing with multiple browser tabs and adjusting the date range.

If you need to control the access to certain goals, you may need to create a profile for certain goals and then give only the people who need access to those goals access to the profile.

A visitor can only convert at each goal once per visit. This is the way it’s always been.

And finally, creating new goals will not modify your historical data, only future data. So all those new goals you’re going to create this week will only track from the day your create them onward.

Share this post:

Filed Under: Analysis, Tracking Tagged With: Analysis, goals, v4

Segmenting Unique Visitors in Google Analytics

October 20, 2009 By Justin Cutroni 18 Comments

Google Analytics now has the ability to EASILY segment Unique Visitors. Some of you may be yawning, but I can hear many, many people saying, “That’s fuc*ing AWESOME!”

Google Analytics can now segment Unique Visitors.

Unique visitors is a critical metric especially in the advertising industry. But it’s also a really hard metric to measure because it take a lot of data processing power. Why? Because of the way unique visitors are identified and measured.

Google Analytics defines a unique visitor as a cookie. For all you nerds it’s the __utma cookie.

Every time a visitor visits a site GA checks for the cookie. If the cookie exists then GA knows the visitor has been to the site before. If the cookie does not exist GA sets the cookie and increases the unique visitor count.

The challenge is that every time you want to view a report that contains unique visitors GA has to literally count all of those cookies collected to find how many are unique. That’s why there was only one GA report with Unique Visitors (Visitors > Visitor Trending >Absolute Unique Visitors report).

But Google figured out some way to effectively count all of the cookies in real time. Now the unique visitors metric that can be added to any custom report.

You can easily add Unique Visitors to any Google Analytics Custom Report.

If you need to segment unique visitors you can simple create a custom report and include this metric.

Here’s an example. Let’s say you’re running a branding campaign for an upcoming movie. You want to measure how many actual people visit your website. You can create a custom report with the campaign dimension and the Unique Visitor metric.

There is one technical limitation. Google Analytics will sample data when a date range for the custom report contains more than 200,000 visits.

Still, I’m completely amazed that they figured out how to make this happen.

Share this post:

Filed Under: Analysis Tagged With: Analysis, v4

Google Analytics Custom Variables Overview

October 20, 2009 By Justin Cutroni 24 Comments

Today Google releases Custom Variables (cv for short) in Google Analytics. This is an evolution of the custom segmentation feature. This post is meant to give you an overview of the feature. We’ll discuss how to use it in a later post.

Like Custom Segmentation, custom variables are a flexible way to add more information to Google Analytics. The big difference is that you can create LOTS of custom variables. How many? In theory you can set an infinite number of custom variables. But GA has some internal limits that keep you to 50,000.

What can we use custom variables for? The possibilities are endless:

Segmenting members from non-members
Segmenting customers from non-customers
Tracking all the campaigns a visitor sees prior to converting
Content categorization
Segmenting visitors based on landing page
Visitor segmentation based on demographic info
Customer segmentation based on order history

Google Analytics Custom Variables are like data decorations!

As my friend Phil likes to say, custom variables are decorations that you hang on your data. Almost like holiday decorations hanging on a tree! This is a really good analogy that I’ll continue in this post.

There are four critical attributes of a custom variable that we must understand in order to use them.

Name and Value

The easiest attributes to understand are Name and Value. The Name of a custom variable is literally the name you give to the variable. Each variable can have many, many values. For example, you could define a variable named ‘Baseball Team’ and then add the values:

Red Sox
Yankees
Phillies
Giants
Angels

This is totally different than the old Custom Segmentation feature. With Custom Segmentation you were limited to one variable (ie one Name) that could contain multiple values. Now you can create multiple variables each of which can have multiple values.

You can view all of your variable names in the new Custom Variables report.

Google Analytics Custom Variables Report

It’s important to note that the name of a variable, plus the value for a variable must be less than 64 characters. Why? The data is sent to Google via a request for an image file. The actual length of the request is limited and Google wants to insure that all of the data makes it to the server.

Scope

The real power of custom variables comes with something called the Scope. Think of scope as the different ‘levels’ of visitor data. When a visitor visits a website Google Analytics collects data at three levels:

Pageview level: This is data associated with each page viewed during a visits. Page level data can change from one page to the next.
Visit level: This is data associated with the visitor’s entire visit. This data can change from one visit to the next. But visit level data is applied to every page within a visit. This data only exists for the CURRENT visits.
Visitor Level: This data is applied to the visitor and every visit and every pageview that the visitor generates. This data persists across all visits that a person creates. How does it persist? Via a cookie.

This means we can set information, ie custom variables, at the page level, the visit level and the visitor level. If we think of custom variables as decorations “hanging” on our data then we could use the following graphic:

GA Custom Variables "hanging" on your data.

So scope is the same as level. Anyone drooling out there?

The ability to control the scope of a custom variable makes this feature extremely flexible. For example, if you want to group all of the content on your site you can add a page level custom variable to every page that identifies the groups that a page belongs to.

If you want to segment visitors by their purchase history you can add visitor level custom variable. The possibilities are truly endless.

Let’s take a look at some of the reporting so you can get a feel for some of the data.

Here’s the Custom Variables report. You’ll notice it looks a lot like the user defined report. This report contains all of the variables that you defined. If you click on a variable you’ll see all of the VALUES for that variable.

So why has google added a scope if we can’t see it in the reports? I’m just going to let you guys speculate. But it’s obviously a critical part of CVs and we should see that data.

Index

The last attribute that we need to discuss is something called the Index. To be honest, it’s really hard to define the index. Basically the index is a technical attribute that helps GA organize all the custom variables on a page.

It’s only used during the implementation, so we’re not going to dig any further in this post.

Speaking of the implementation, you’ve probably noticed that I haven’t talked much about the implementation. To be honest, we’re still playing with CVs. Obviously this data comes from JavaScript. So you have to do some coding to get this data.

But I’m going to hold off on the implementation talk until later. Implementation involves another concept called the Index which is, to be honest, vague and confusing.

Share this post:

Filed Under: Analysis, Tracking Tagged With: Analysis, custom variables, v4

Tracking Zero Result Searches in Google Analytics

September 8, 2009 By Justin Cutroni 25 Comments

I <3 Google Analytics Site Search reports. There’s amazingly actionable data in those reports. But they’re missing one vital piece of information: searches that don’t produce any results.

Why is this important? Don’t you want to know when visitors search and don’t get any results? Zero result searches can help your identify missing content on your site or a problem with your site search engine.

Many search solutions will provide this information for you. For example, I use Search Meter for WordPress and it shows me which search queries generate zero results. But I thought it would be interesting to add this data to Google Analytics. That way all my site search information would be in one place.

Unfortunately there is no easy way to add this data to GA. You need to do some programming to collect the data. So this post is really meant for those folks with programming resources AND for those developers that maintain GA plugins. Like my buddy Joost, who has a great GA plugin for WordPress.

If you’re interested in the data and analysis, skip to the bottom of this post.

Conceptual Overview

Our goal with this hack is to modify site search data in two ways. First, we’re going to put all search queries with zero results in a category. This will allow us to use the Search Categories report to easily find all the search terms that yielded zero results.

Second, we’ll modify the actual search terms to indicate that a term yielded zero results. This will make it easy to scan a list of all the search terms and identify which generated no results.

Before we get into the implementation, a big THANK YOU to Charles Miller, one of the lead consultants here. He wrote the JavaScript below. Thanks Charles.

Step 1: Identify No Result Search

The first step is to identify a zero results search page. Most websites have the same search results page regardless of the number of results. You need to identify some something that differentiates a zero results search page from a non-zero results search page.

This must be done programatically and is the hardest part of the implementation.

For example, a zero results search page on this blog has the text “No posts found. Try a different search?”

I can create code (or more specifically, Charles can create code) to look for the text “No posts found. Try a different search?” If the code finds this text in the page then I can identify that the visitor’s search yielded zero results and than I can send the data sent to GA. Here’s the code that I’m using on this blog:

var content = document.getElementById('content');
if (content.innerHTML.search('No posts found.')) {

The code looks for a section of the page called ‘content’ and then searches that section for the phrase ‘No posts found.’. If ‘no posts found.’ is found (oh, the irony!) then we will modify the data sent to GA.

Important! The way you detect a zero result search page may be different. It’s VERY difficult to create an example that will work for everyone. Take this as a conceptual overview.

Step 2: Tweak GA Tracking Code

Once we know what differentiates a zero results search page we can add some code that tweaks the data. Remember, we want to modify the data in two ways: 1. by placing it in a special search category and 2. by modifying the search term to indicate it did not yield any results.

To create the category all we need to do is add an extra query string parameter to the URL.

To manipulate the search term we need to split apart the page URL and then put it back together with the phrase no-results.

Here’s the complete code.

<script type='text/javascript'>
var pageTracker = _gat._getTracker("UA-XXXXXX-1");
var content = document.getElementById('content');
if (content.innerHTML.search('No posts found.')) {
     // These lines get the search data from the URL and
     //  deconstruct the URL into parts
     var sn = "s";
     var sr = new RegExp(sn+"=[^\&]+"),
      p = document.location.pathname,
      s = document.location.search,
      sm = s.match(sr).toString(),
      srs = sm.split("="),
      // The next line is where we add the category and add
      // the phrase no-results to the search term.
      sre = sm.replace(sr,srs[0]+"=no-results:
 "+srs[1]+"&cat=no-results"),
      sf = s.replace(sr,sre);
      // Send the data to Google as a Pageview
      pageTracker._trackPageview(p+sf);
} else {
      // If this is a regular page on the site, use the standard GA code.
      pageTracker._trackPageview();
}
</script>

The code starts with the section that identifies a zero result search page.

Then we deconstruct the URL to identify the search term. Finally we add the category named ‘no_results’ and the phrase ‘no-results’ to the search term.

If the code does NOT find the term ‘No posts found.’ then a pageview is created as normal.

That’s it for the coding part (thank goodness!)

Step 3: Configure Site Search Settings

The last step is to add the new category parameter to the Site Search settings so GA can identify the no-results search category. This is easy, it’s in the profile setting section of Google Analytics.

I also like to set the ‘Strip Query Parameter’ to YES. This removes the category parameter after site search is done processing and normalizes your pageview data.

That’s it for the configuration! We’re cleared for insight-hunting!

Analyzing The Data

When a visitor performs a search that yields zero results the search term will be placed in a category named ‘no_results’. To find this data navigate to the Content>Site Search>Categories Report:

Immediately you’ll be able to see what percentage of your searches yield zero results. Hopefully it’s very low! Want to see if this impacts conversions or revenue? Click the Goals or Ecommerce tab to check the conversion rate:

This is a bad picture, but you get the point.

Next you can click on the no-results line in the data and see exactly which search terms yielded zero results.

This is super-actionable data. Now you know where you may be missing content or if your site search engine might be broken. You should be asking yourself, “Why are there no results for these terms? Is there missing content or is there a problem with my site search engine?”

You’ll also notice that the search terms now have ‘no-results’ in them. This provides a lot of flexibility for view the search data other ways. Example, let’s use the Search Terms report:

Here we can see the search terms ranked by searches. What percent of your top 10, 20 or 50 are no-result searches? How is that impacting your bottom line?

This is just the start. You can use other metrics, like %Search Exists to understand if visitors who receive zero results refine their search or exit.

While this is not the easiest thing to configure, I hope you see the value of the data. More so, I hope that all those folks that maintain plugins add this type of feature to their GA plugins. Joost, you listening!?

Share this post:

Filed Under: Analysis, Tips, Tracking Tagged With: enhancements, site search

Segmentation Options in Google Analytics

April 21, 2009 By Justin Cutroni 25 Comments

As web analysts we live and die by segmentation. Without the ability to segment traffic we can not isolate which segments are producing and which need improvement.

Google Analytics offers many different ways to segment data. Each has pros and cons but there is always a way to get the data you need…. well, almost always.

There are 6 different ways to segment data in Google Analytics:

1. Using certain reports
2. Dimension drop down
3. Report filters
4. Advanced segmentation
5. Custom reports
6. Profile filters

Bet you didn’t think there were SIX ways to segment data. :)

Using Certain Reports

Ok, you may think this form of segmentation is lame, but it’s not!

Many reports in Google Analytics are segmented by some default piece of information. There’s nothing for you to do.

For example, the Browsers report segments your data based on the different web browsers that visitors use to access your site. Google Analytics automatically identifies this information when collecting visitor data.

Google Analytics Browsers Report

Other segments that are automatically include in Google Analytics include:

* Visitor type (new and returning)
* Geographic information
* Operating system (and many other “nerd” segments)

Marketing segments are not AUTOMATICALLY segmented. You need to configure Google Analytics to track campaigns (i.e. link tagging) in order to get correct traffic source segmentation.

Pros:
* Easy

Cons:
* You better make sure you’ve got your campaigns tagged correctly ;)

Dimension Drop Down

Many reports have a dimension drop down that allows for segmentation right in the report. This is a handy way to quickly drill down into a piece of data.

For example, let’s say I want to see the most popular landing pages in a particular state. I can navigate to the state in the Visitors > Map Overlay > Regions report, click on the state I’m interested in, and then choose Landing Page from the Dimension drop down.

Google Analytics Dimension Drop Down

You can see in the image above that you can segment based on campaign information, some technical information and some visitor information (visitor type, language).

Overall, this is a good way to go when you’re drilling down and want to segment a single data point by some dimension.

Pros:
* Quick for one-off segmentation

Cons:
* Can trigger sampling
* Limited number of dimensions and no metrics
* A pain if you need to segment a lot of things, like top landing pages for every US state

Report Filters

Bet you don’t think of filtering as segmenting, but it is!

Any report displaying tabular data has a filter tool at the bottom of the data. This let’s you quickly view data that matches, or does not match, some condition. The condition is the pattern, or regular expression, that you enter into the filter. Using a regular expression you can add lists to the filter.

Here’s an example. Suppose I want to quickly view traffic coming from the Pacific sales region. I can apply the following filter to the Visitors > Map Overlay > Regions report:

California|Oregon|Washington

[The above is a regular expression matching California OR Oregon OR Washington]

A filter Map Overlay report in Google Analytics

Notice that the Scorecard (the top row of data in the table) indicates how our segment, i.e. the data that matches our filter, compares to the overall site? We can now compare the Pacific sales region to the entire site.

And here’s a neat trick, if you add the filtered report to your dashboard the filter will persist in your dashboard widget. I call it a sticky filter.

Pros:
* Quick and relatively easy
* Can be applied to historical data
* Will not trigger sampling

Cons:
* Restricted to one report and the data in that report
* You should know some basic regular expressions
* Can not be shared easily

Advanced Segments

There has been a lot of conversation over the last few months about Advanced segments and rightly so. This analysis tool is really powerful and let’s you slice the data many different ways using different dimensions and metrics. Want to see all visits that generated more than $100, coming from paid search and occurring after 8 AM? No problem with an advanced segment.

An Advanced Segment in Google Analytics

But there are some downsides. First, sampling. Because Advanced segments re-process data in real time there is a sampling algorithm applied to minimize the load on Google’s servers.

You can’t segment more than 200k visits. If sampling is applied you’ll see a confidence interval next to your data.

Sampling accuracy in Google Analytics

Again, the problem is that small segments of data will be really inaccurate when the sampling algorithm is applied. There is no way to disable sampling.

The most common ways to get around sampling are segmenting using profile filters or potentially a report filter. It really depends on the exact situation.

The second issue is that not all reports can be segmented. Due to the segmentation technology certain reports can not be segmented, like the Absolute Unique Visitors report and the Funnel visualization report. Those reports can only be segmented with profile filter (see below).

Pros:
* Can be applied to historical data
* LOTS of flexibility, can segment based on a huge number of dimensions and metrics using different combination of both

Cons:
* Sampling will be applied if trying to segment more than 200,000 visits
* Not all reports can be segmented
* Specific to your username, can not be shared with other users

Custom Reports

Another beta feature that can be used for segmentation is the Custom Reporting tool. This tool is more than just pretty reports. It allows you to create 5 levels of segmentation in a report.

In a previous post I talked about segmenting campaigns by time of day to better understand day parting.

We could take that example one step further by adding geographic region to the report. The result would be a report that has Campaigns data that could be segmented by time of day and then by geographic location.

Multiple levels of segmentation in a GA Custom Report

The problem is that not all dimensions can be used together. The reason is that only certain metrics are related in the Google Analytics data architecture. You can find a complete list of combinations in the GA support docs.

Pros:
* 5 levels of segmentation
* Advanced segments can be applied to a custom report
* Can be shared using automated email feature

Cons:
* Can only segment using dimensions, not metrics
* Limited number of dimension combinations
* Can only drill into one data point at a time

Filtered Profiles

Filtered profiles are the nuclear bomb of segmentation. They are permanent, segment every report in a profile, and can easily be shared.

In case you’re not familiar with filtered profiles, you can include and exclude data from a profile using a filter. Google applies the filter during data processing, thus segmenting the data.

Google Analytics profile flters

Once the data has been processed it can never be changed. This means that you can filter historical data AND if you mess up the configuration of a filter you could have really crappy data.

Another issue with filtered profiles is not all data can be filtered. For example, transactional data is different than pageview data. This can cause some funky information in campaign reports and commerce reports.

If you need to filer pageview data then you’ll need to filter your commerce data with different include or exclude filters. Also note that Event data can not be filtered.

But, on the up side, you can use filters to segment things like Absolute Unique Visitors and the Funnel Visualization report. Neither can be done with an Advanced Segment.

Absolute Unique Visitors in Google Analytics

Pros:
* Segmentation of every report in Google Analytics
* Can control access by assigning users to filtered profiles

Cons:
* Only effective from date of implementation forward
* Limited number of dimensions
* Issues with filtering other types of data, like events and transactions

I hope this inspired you to come implement different segments using different techniques. As I said in the beginning, there are a lot of ways to slice data in Google Analytics. Find the technique to suit your needs and start segmenting!

Share this post:

Filed Under: Analysis, Tips, Web Analytics Tagged With: advanced segmetnation, custom reports, filters, segmentation

Step 1: Create a New Profile

Step 2: Tag your Internal Campaigns

Step 3: Configure Site Search Settings

The Reports

Goal Sets

Goal Types

Time Based Goals

Pageview Based Goals

URL Destination Goals

Funnels

Odds and Ends

Name and Value

Scope

Index

Conceptual Overview

Step 1: Identify No Result Search

Step 2: Tweak GA Tracking Code

Step 3: Configure Site Search Settings

Analyzing The Data

Using Certain Reports

Dimension Drop Down

Report Filters

Advanced Segments

Custom Reports

Filtered Profiles

Subscribe & Connect

GET POSTS BY EMAIL

My E-book

Recent Posts

Categories

Tags

tweet tweet tweet [my tweets]

Archives

Blogroll