Segmentation in Google Analytics - Part I

Introduction

Segmenting traffic in Google Analytics (or any other web analytics pacakge) is key for any analyst who is looking to get the most out of their data.  I have seen far too many businesses look at their analytics data in the aggregrate, without taking advantage of multiple profiles, advanced segments, or advanced filters.  The “slicing and dicing” of data that can be done in Google Analytics can really provide a tremendous amount of insight into one’s online marketing efforts - be they SEO, PPC, or Social Media.

The following a brief series of blog posts highlight some real client cases that provide you some tips about how to segment your traffic in different ways (and “why” you should be doing it).

Segmentation & SEO

I was doing an SEO review for a client recently who was interested in getting the most ‘bang for their buck’ from there SEO efforts.  In particular, they were interested in knowing which terms should they be spending their time (=money) working on internally and outsourcing to SEO firms.

So my question was, “how have your SEO efforts been performing to date?”

Their answer: ”Pretty good.  Traffic seems to be going up.”

Organic Traffic

This was true.  Their traffic was going up.  But which traffic?  Branded searches?  Non-branded searches?  I knew that they were doing some marketing recently and that there has been some good buzz about their brand name.  The quickest way to get a sneak peek at branded and non-branded search terms, is to filter by keyword at the bottom of the page.

Where to type in your filter

The result:

Branded Traffic

While looking at Branded Traffic can provide a nice measurement of one’s brand awareness and serve a signal to measure the impact of marketing efforts (the spike in branded search was due to some PR buzz) - it certainly doesn’t help our desire to find the “bang for your buck” SEO opportunities.

So I changed the filter from “containing” to “excluding” and got:

Non-Branded Traffic

We quickly see that their SEO efforts have been driving some more traffic to their site.  Filtering is nice because it can provide you with the view you want very quickly.  Obviously we’re not done here, but we do know that before delving a bit deeper into the search for “bang for your buck” SEO opportunities, that it makes sense to segment organic traffic to look at branded and non-branded keywords.

Let’s quickly make two Advanced Segments for these different types of keywords.

We now have a much clearer look into this client’s Organic Traffic segment.  As to be expected, branded traffic has a lot more return visitors (% New Visits) and a much lower bounce rate, which in turn effects the other site usage metrics such as Avg. Time on Site and Pages/Visit.  While brand awareness is very important for this business, by spending time looking only at non-branded organic traffic, efforts can now be spent on driving new customers to the site.

Comparison of Branded and NonBranded Organic Traffic

We now turn our attention to e-commerce.  Since our goal is to get the most “bang for our buck,” we need to find those keywords which convert.  I chose to use a simple advanced filter where, based upon the number of transactions and traffic that this client has received, I set the number of transactions to greater than zero and then sorted by descending by the number of transactions.

Organic non-branded keywords - sorted by most transactions

Now that we have a sense of which keyword organic keywords are the best performing on the site, my next step is to export all the keywords into a spreadsheet and then copy and paste all of the keywords into the Adwords Keyword Tool.  This tool will provide us with information about the search volume for these keywords so that we can determine which converting keywords have the most potential.  Once we know which keywords have the most search volume and we believe they will convert, we can work out an SEO strategy based on inbound links and on-page optimizations that will target those keywords.

Stay Tuned for Part II….

I have to take a moment and give a shout out to a top-notch Google Employee that I have worked with for the past almost two years.  Kevin, you’ve been a most excellent ‘Relationship Manager’ as I suppose your official title was.  Thank you so much for your many hours on the phone with me as we worked on the account together.  I wish you the best of luck in you future roles at Google.

I recently started working on PPC and Analytics for a new launched website.  In this case, it is an Optometrist in Lincoln, NE.  I’m particularly interested in knowing if anyone has any bounce rate benchmark information for medical practice sites.  Please comment below if yes.  It seems to me that many private practices could greatly benefit from a knowledge Analytics expert.  I’ll keep everyone in the loop as to how this newly launched site is panning out.  I’ll even probably share some groovy charts.

One report that I find extremely useful in the Goals Section is the Conversion Rate Report.  In a moment, I’ll explain how a quick export of data from the Conversion Rate Report can provide valuable insight into one’s abandonment rate.  But first, a quick intro into conversion rates, abandonment rates, and conversion funnels.  In general, I assume that my readership is relatively erudite in Google Analytics.  So if this next paragraph is below your level of expertise, feel free to skip ahead.  That said, I believe it is worthwhile for all us out there who are trying to be successful on the Internet to get back to the basics and remember the fundamentals.

Fundamental #1 –>  It’s all about conversions.

Really.  It is.  Especially when it comes to e-commerce.  There are soooo many people about there who are still interested in how many people come to their site.  But if they aren’t taking the desired action(s) that you would want them to, then most likely you’re wasting time and money.

Take a step back.  Take a deep breath.  And say to yourself, “what do I want people to do when they come to my website.”  The answer(s) should be easy.  Lead generation?  (Filling out a form).  Signing up for an email list?  Online purchase?

Once you define your goal and properly configure it in the settings section, you’re ready for the next step.

But before we get there….

Fundamental #2 –>  Lowering Abandonment Rate is the best way to increase conversions.

For most of the online conversions we can think of, there is usually a multistage process that the user needs to go through in order to complete the conversion.  This is called the conversion funnel.  Whether it is filling out a form, or completing an online purchase, if the user is on one page and then needs to make a click on a button to submit or proceed, they are in the funnel.  If a user is in the funnel, but they don’t make it to the end of the process, then they are ‘abandoned’ the funnel.  In many cases, a potential customer will put an item in their shopping cart but won’t check out.  This is called shopping cart abandonment.  This is also called leaving money on the table; or imagine yourself at the grocery store with a whole cart of food and the you just leave the cart in the middle of aisle 6 and walk out of the store without buying anything.  Since you don’t want people to walk out of your store, one of the best things to do is to get less people to leave when they have good in your cart.

This article is not about ways to lower cart abandonment.   There are a lot of good tools out there (read: Google Website Optimizer) that can help with testing different cart scenarios and designs.

The main tip that I’m presenting here is a unique way to look at abandonment rate.  By noticing trends in abandonment rate you will likely find a way to leverage this information to make some decisions that will pay off (like adjusting your PPC bids).

Ok… enough chatter.  Here’s the tip:

As I mentioned in a previous post, it is very worthwhile to set up goals for different stages of your funnel.  This may require setting up an additional profile to do just this, but trust me, it’s worth it.

In this example, we’re going to look at those people who put something in their cart and those who complete their purchase.

Simple math –> divide the number of people who complete their purchase with the number of people who put something in their cart, subtract from 1, and you have your abandoment rate, represented as a percentage.

Abandonment Rate = 1 – (purchase/cart).

In the conversion rate report under the goals section a really nice feature is the ability to look at conversion rates at different times of the day.

Conversion Rate by Time of Day

Remember that simple math from above?  So just export your two reports with conversion rate for time of day at the shopping cart and completed purchase levels of your funnel and you can calculate your abandonment rate by time of day using a spreadsheet.

Abandonment Rate by Time of Day

An action item for this client would be to lower PPC bids during off hours and boost them during midday hours.

Questions and Comments are welcome.

]]> http://www.analytics-ninja.com/2009/10/conversion-funnel-analysis-abandonment-rate-by-time-of-day/feed/ 0 http://www.analytics-ninja.com/2009/09/google-analytics-goals-bounce-rate/ http://www.analytics-ninja.com/2009/09/google-analytics-goals-bounce-rate/#comments Mon, 07 Sep 2009 12:19:37 +0000 admin http://www.analytics-ninja.com/?p=69

Proper configuration of Goals in Google Analytics is a critical to the ability to make good business decisions.  Especially for those of you out there spending money on PPC, in order to make the best return on investment possible (and stop wasting your money), properly set up goals are a must.

In my humble opinion, Conversion Rate and Bounce Rate are two of the most important key performance indicators (KPIs).  As you can see below, when I configured Goal Conversions for a client, I was able to make the types of decisions necessary in their PPC account to drive conversions up and bounce rate down.

BEFORE –> 50.19% Bounce Rate | No Conversion Tracking.

Bounce Rate Before

AFTER –>38.73% Bounce Rate | 2.28% Conversion Rate

Bounce Rate After

As you can see, I was able to take the bounce rate down almost 12 percentage points.  This is significantly lower than the benchmark for sites of similar size.  Check out the stats for last week’s traffic.

Bounce Rate Benchmark

The major improvements in online performance were mostly do to looking at which campaigns were working and which weren’t.  While bounce rate is a very important metric, I would not have felt comfortable making such bold changes to the PPC spend without knowing the conversion rate.

Comparing Campaigns | Conversion Rate

Where would you want to put your money?

Another extremely useful place to look at in Google Analytics is the Map Overlay.  The Map Overlay allows you to look at Bounce Rate and Conversion Rate by location (of course, in addition to other valuable metrics).

This particular client was spending close to 25% of their PPC budget for ads in the UK.

Conversion Rate by Country

Yuch!  But before I decided to pause the ads in England, I applied an Advanced Segment that allows me to see the performance of all visitors who didn’t bounce.  Just wanted to make sure that I knew about the behavior of the Brits who did stay on the site.

Conversion Rate by Country | No Bounce

Also pretty bad… Guess who isn’t advertising in England anymore?

Thankfully, looking at these metrics smartly has allowed this client to save significant money and improve their ROI.

In any case, I found the relationship between conversion rate and bounce rate to be interesting.  I hope you did too.

Until next time….

Ninjafully yours,

A.N.

Understanding your conversion funnel is one of the keys to finding the best ways to improve your conversion rate.  The first thing to do is to identify which part of your funnel needs fixing.  For example, are people abandoning in the shopping cart?  Somewhere else in the checkout process?

conversion funnel

The challenge that I’ve come across is answering questions like, “Who is abandoning my cart?  Which keywords are they searching on?  Which traffic source are they from?” and on and on…

The Supreme Ninja already told us that by using Advanced Segmentation we can all be Analysis Ninjas.  Well, here is my tip for super-charging your conversion funnel analysis.

Really, it’s quite simple.  Just create an Advanced Segment for each page in your goal funnel.  (Dimension > Content > Page = Step in Funnel).  In the example above, the cart and form pages of the online purchase funnel are the two areas where lots of people are dropping out.

Here’s an example of creating a segment for all visitors who visited the cart page:

advanced segment of shopping cart visitors

The results are quite powerful.  Let’s take a look at All Traffic Sources.  We can now export this data and calculate the abandonment rates for each of our traffic sources.

advanced segments of conversion funnel

This can now be applied to the Map Overlay (which locations have higher or lower abandonment rates), the Content Section (which pages have higher or lower abandonment rates), Keywords, Campaigns… use your imagination.

Let me know what you think!

Best wishes,

Your friendly neighborhood analytics ninja

This Google Analytics Ninja seems to have saved the day and spotted a problem with code implementation for a friend’s WordPress blog.  (Well, a friend of a friend, but I guess it is about the same thing).

On July 20th, I was approached with the all too common, “I believe that I”ve configured Google Analytics properly, but it isn’t working!” problem.

I took a look at their source code, and I spotted an extra space.

Google Analytics Isn't Tracking - problem

I thought to myself, “hmm… how did that space get there?”  In the WordPress backend, I confirmed that indeed a space was the culprit.

Google Analytics Isn't Tracking - Culprit

I don’t know about you, but I don’t think I would have caught that problem just by looking at the Analytics for WordPress plugin configuration page.

Shortly thereafter, we were able to see GA tracking properly.

Proof of Success

Action Item:  Check your source code for any unwanted spaces.

Extra note: this can be the culprit for e-commerce tracking issues as well.

There are a lot of things that bother me about Yahoo Search Marketing.  A LOT.  If anyone from Yahoo out there reads this blog, please pass on this piece of advice to your supervisors –> copy Google.  DO WHAT THEY DO.

*Sigh*  I’ll vent more frustration about Yahoo later.  There does happen to be, however, one or two areas that Yahoo is superior to Adwords, imvho.  Today’s post is about one of them:  Tagging.

For the purposes of this post, I assume that everyone reading this blog knows how to tag their links.  If not, please read up here.  And use the URL builder.

Here is the tag that I put on all of my Yahoo PPC destination URLs.

?utm_source=yahoo&utm_medium=cpc&utm_campaign=PUT-THE-CAMPAIGN-NAME-HERE%20{YSMMTC}&utm_term={YSMKEY}&utm_content={YSMRAW}

Another option is to leave the match type off of the name of your campaign.  I have found it helpful to leave it in, like the first example.

?utm_source=yahoo&utm_medium=cpc&utm_campaign=PUT-THE-CAMPAIGN-NAME-HERE&utm_term={YSMKEY}&utm_content={YSMRAW}

The utm_term={YSMKEY} tag populates the Keyword dimension with the keyword that you are bidding on, and the utm_content={YSMRAW} populates the Ad Content dimension with the raw query.

Make sure that you have tracking urls turned on in the Yahoo backend.

The result in Google Analytics is the ability to drill down on a keyword in Yahoo and see all of the raw queries which triggered the keyword.  In other words, a search query report (which, not surprisingly Yahoo doesn’t even offer in their UI, you have to request it from the support team.  argh! )

Yahoo Search Query Report

IMPORTANT:  Make sure to exclude the URL parameters from your Analytics reports.  Otherwise it messes up your Content Report, landing page analytics, etc.

exclude URL parameters

This is one of my favorite advanced segmentations. I often times find myself frustrated by the fact that keyword match types are all mashed together and reported as one keyword. Indeed, as far as I am concerned, they are three different keywords. For example, in Adwords

[podiums]

“podiums”

podiums

will have different CPCs, different average positions, different conversion rates, but in Google Analytics I only see

podiums

With one bounce rate, one conversion rate, one average pages per visit, etc.

Since I’m bidding and paying different amounts for different match types, I treat [podiums] differently than “podiums”. Wouldn’t it be nice if I could natively segment this in Analytics?

Well… unfortunately you can’t.

Keyword Positions

But here is my method to do exactly that.

Step One: Embedded Match

Separate all of your keywords by match type into unique adgroups. Then use embedded match to guarantee that each search query triggers the right keyword match type. Using embedded match isn’t fully necessary here, but it is a good idea anyway.

What is necessary is that the adgroups are clearly named so that they can be segmented later.

Embedded Match Adgroup Names

Step Two: Create Advanced Segments

Segment by Match Type

Choose Adgroup as your Dimension. Choose contains and then the naming convention of your adgroup. In the example above, I have named all adgroups that contain phrase match keywords with the word “phrase” as a part of the adgroup name.

Now, when I look at any of my keyword level reporting in Analytics, I can see how different keyword perform by match type.

What it looks like when you segment by match type

Questions and Comments are welcome.

Enjoy,

Analytics Ninja

Welcome to my Analytics Ninja Blog:

Analytics Ninja

Yes… Yes… I know that there are plenty of other Web Analytics Ninjas out there. I have a lot of respect for many of my colleagues who are continually rocking the field of Web Analytics with their Analytics Awesomeness. That said, with a great sense of humility I do feel entitled to share the designation of Analytics Ninja with my fellow ninjas who are out there.

I hope that any readers of this blog will find the musings of this ninja to be informative, worthwhile, and enjoyable.