The practical guide to Google Analytics Advanced Segments

This week I’m aiming to spend 3 hours on 12 different free traffic generation techniques but before I do, I want to make sure I can track the results properly. And I’m going to do that by using Google Analytics Advanced Segments.

I do love Google Analytics and the standard reporting is great however if you want to break down your traffic by marketing strategies the standard reporting won’t really cut it. So let’s start there.

Problems with standard reporting in Analytics

So there are a few problems with just looking at the standard analytics reports.

Campaigns don’t cover everything

Campaigns are great, here’s a list of my campaigns that I’m playing around with on my simplified business reporting app startup Informly.

Campaigns are great but they don’t cover any marketing where you can’t control the URL that people link to (which could potentially be a lot).

So when you structure your marketing you might want to try a lot of different things like I am doing and you want to compare results side by side. You can use trackable links / campaigns for initiatives where you have full control over the URL that people go to (email marketing / adwords / forum signatures etc) but this won’t cover all of your marketing (i.e. when people tweet your articles, or mention you in forums or when you are listed on other sites etc).

There are too many sources

Sources are great too but there are too many of them to take meaningful high level lessons from.

Sources are great but even before I do much active marketing there are already 70 different sources of traffic to my website.

For example one of my marketing methods is to get listed in the partner directories for the people I have integrated with Informly (like MailChimp, Xero etc). This is a marketing strategy I’m calling ‘Partner integration’ and I want to know the results of this strategy as a whole. To do this with sources I’d have to manually add up all of the traffic from all of my partners.

Overstating of natural search traffic

Another big one is how much analytics overstates natural search traffic due to how many people go Google and put in brand keywords. In my case for example natural SEO is the biggest referrer to the Informly site (16%). I might get the idea from this that the articles I’m writing on the site are having an impact. Not true.

When you take out the people going into Google and putting in ‘informly’ who really should be treated as direct visitors the result goes down to 12% and importantly conversions go down a lot as well.

So I need to find a more accurate reflection of natural search traffic and direct traffic to help me make better decisions around spending my time on efforts designed to bring in natural search traffic (i.e. onsite content and offsite link building).

The answer Google Analytics Advanced Segments

Fear not, Google has the answers. While standard reporting can’t do what we want, we can achieve it with advanced segments – specifically by defining our own custom segments in the advanced segments area.

What are advanced segments?

Advanced segments allow you to view and compare your traffic by breaking it up into more flexible categories or segments. There are some provided by Google Analytics but the real power is in defining your own segments.

Uses for custom segments

Here are a few ways I am using custom segments to analyse the traffic to Informly.

  • I have a ‘Direct’ custom segment that includes all direct traffic as well as anyone who goes into Google and puts in brand keywords. This to me is a more accurate way to measure direct traffic.
  • I’ve got a category for Social media that includes all of the main social media sites but doesn’t include specific trackable marketing I’m doing on those sites (see below).
  • I’ve got a Twitter auto follower strategy I’m working on and for that (and a number of other strategies) I’ve got a trackable link set up which I have added as a custom segment. Yes you can just use campaigns for this but to compare side by side with other strategies that can’t be captured as campaigns it’s useful to create a stand alone segment for it.
  • I’ve got one set up for ‘Natural product coverage’ to me, making the product better is my most important marketing effort and when I continue to do that hopefully I’ll get coverage in other websites and blogs and I’ll add all of those to this custom segment when they cover the app (a few places have covered it already).
  • I also see my Web Domination podcast as a traffic strategy for the app so I’ve created a custom segment just for traffic that comes via that.
  • I’m going to look at doing some videos so I’ve created one for video sites that wil include Vimeo and Youtube.

How to set up custom segments

I was able to set up all of my segments in around 2 hours, here is the step by step process.

Step 1 – Open Google Analytics Advanced Segments

So first off open Analytics, make sure ‘Standard Reporting’ is selected up the top and click on any of the pages on the left to view a standard chart (let’s just choose the Overview chart). Then click on the ‘Advanced Segments’ button to open the advanced segments area.

The advanced segments button is available above all of the standard reporting charts in Analytics.

 Step 2 – Create a new custom segment

On the Advanced Segments screen you will see the inbuilt advanced segments on the left and the custom segments on the right. Yours will be blank but once you add a few it will look like the screenshot below. To add a new one click on ‘New Custom Segment’.

Here are my custom segments on the right – you can use this Advanced Segments screen to add new segments and view / edit your existing ones.

Step 3 – Create your segment using Or & And statements

This is where things can get a little tricky. You need to think about what needs to be included in your segment and what needs to be excluded to provide actionable data on your marketing efforts. Let’s take the example of finding a more realistic measure of natural search traffic.

In this case we want to:

  • include any traffic that came from the major search engines (Google / Yahoo / Bing), and
  • make sure only organic traffic is included not paid traffic, and
  • exclude anyone who was searching for our brand ‘Informly’.

Check out this screenshot for the settings we use for this.

Custom segments including all organic search traffic minus people searching for ‘brand’ keywords.

Here is a description of the rules I have set up according to the statements made above.

  1. I’m including everyone from the source that mentions Google, Yahoo or Bing. Note you could add 3 different OR statements for this – i.e. include traffic from Google, or Yahoo, or Bing. But the ‘Match Regular Expressions’ operator enables you to do it all from within the 1 statement by separating the items with a | sign.
  2. I’m only including traffic that is is from the Medium ‘Organic’. Because I’m using an ‘And’ statement here it will effectively exclude any traffic that comes from any other medium – i.e. paid search traffic.
  3. I’m excluding any traffic that comes from my brand keywords again using a ‘Matching regular expression’ operator to include all of the keywords within the 1 statement (this expression includes my exact domain, my brand name my own name etc).

Continue this process for all of the components that make up your marketing strategy and you’ll then be able to view charts from different segments.

Reporting on different segments

So the final step is the good bit. Once it’s all set up you can then start to analyse your traffic from your various strategies. You can either do this within Google Analytics or if you use Informly we have 2 very simple charts that give you all of the important data in the one place. We’ll start with reporting in Analytics.

All you have to do is visit any of your favourite standard reports and click the Advanced Segments button mentioned above and select up to 4 segments to compare and click ‘Apply’.

Here are 2 examples of how I am using it.

Traffic Overview report

Looking at different overall traffic measures for 4 different marketing strategies side by side.

So the above chart (slightly modified so I could you show you more on the page) is looking at some overall traffic figures from 4 different marketing strategies:

  1. Direct (not really a strategy)
  2. long tail SEO traffic (I’m using this to measure how effective my onpage blog articles are)
  3. a Twitter Auto follow strategy I’m working on (will write more about this another time)
  4. Partner integration (this is where I see out partners with directories that will list us in))

There are some interesting learnings from this including:

  1. Direct traffic is by far our biggest referrer, the natural search traffic is a lot less than otherwise stated.
  2. Traffic from the Twitter auto follower strategy and partner strategy are considerably more engaged on average viewing 1 more page per visit and spending around a minute longer on the site. Definitely worth exploring these further.
  3. The bounce rate on longtail SEO traffic is the worst (meaning people are coming, reading one article and leaving) and on the partner strategy is by far the best (meaning partner traffic is much more targeted traffic or warmer leads).

Conversions

Conversion charts give me useful information on which strategies are converting visitors to users.

The main thing I am interested in is who is signing up to use the Informly app.

If I use the same segments as above and look at the conversion chart I see:

  1. Direct traffic is converting by far the most visitors although the conversion rate is quite low (this is probably because a lot of current users would go to the site just to sign in) but it still tells me that most people are hearing about the app outside anywhere that I can track.
  2. While long tail SEO is bringing a lot of traffic it’s bringing hardly any customers at all with a terrible conversion rate. This could be a content issue or perhaps it’s just not an effective strategy to get users.
  3. The twitter auto follow and partner strategy have an exceptionally high conversion rate of over 5% which means I’ll be looking at doing more of this.

Let’s make it simple

The custom segment conversion chart in Informly gives me a clearer picture of the traffic quality by ordering my segments by conversion.

One of the issues above is you can’t report on more than 4 segments at once. With Informly’s custom segment reports you can view your top 10 strategies on the 1 simple chart.

We offer 2 charts, one will list your top segments by traffic source and the second (pictured on the right) will list them in order of conversions. This way you can quickly identify your top sources of customers.

More info

Here are a couple of other great articles I read while learning about custom segments that you may find useful.

Kiss Metrics article - This is an excellent place to start and spells everything out and also provides some great lists of statements you might want to include for various marketing strategies.

One thing missed in this article however is the ability to use ‘Matches regular expression’ to group sites together for example I have one for ‘natural search traffic’ that uses this feature to include Yahoo, Bing and Google.

This search engine journal article goes into this in more detail and also provides some more SEO related information (note the version of Analytics has changed a bit since then).

I hope you have found this useful please feel free to comment if you have any questions.

About

Dan Norris is a co-founder at WP Curve and a passionate entrepreneur with an obsession for content marketing.

  • http://www.insitedesignlab.com/blog insitedesignlab

    This is a great post, I’ve been really trying to better understand my analytics data lately. Thanks!

    • http://inform.ly/ Dan Norris

      Hey Jon thanks. I’m pretty excited about custom segments we are working on a neat chart for Informly that will show your visits based on up to 10 custom segments. And also one that will show your conversions for up to 10 segments. This isn’t even something you can do in Analytics so I’m pumped about the ability to present that information to people – it’s proving very useful for my business.

  • http://twitter.com/jingpingji JP

    Hey Dan! I was reading Perry Marshall’s Definitive Guide to Google Adwords and then was researching about “advanced segments” and there you were on the first page of Google! =) Very helpful post. I think I’ll be making good use of advanced segments after reading this.

    • http://inform.ly/ Dan Norris

      Ha! nice.

  • http://www.blazewebstudio.co.za/ geoffreygordon

    I stumbled onto this from your guest post on think traffic, which i found on social media examiner, i definitely need to look at my metrics on google analytics and do some tweaking. PLus you gave me a better understanding as to what i can do with google analytics thanks

    • http://inform.ly/ Dan Norris

      Hi Geoffrey awesome I’m glad it helped! I’m building in an advanced segments chart into Informly so users can see a chart with their top 10 segments with traffic and conversions. This will hopefully be in the version we launch in 4-5 weeks.

  • Josh

    Hey Dan,

    Nice post.. I had a quick question about the advance segment. Is there a way you can differentiate them according to the profile because once you save the new custom segment it gets applied to all the profiles?

    Thanks
    Josh

    • http://inform.ly/ Dan Norris

      Hey Josh I’m not sure about that. My understanding was they will apply across all of your profiles. But I am only using them for my main webcontrolroom.com domain so I haven’t played with the profiles. Hope you get it sorted mate and thanks for stopping by.

  • http://twitter.com/7th_st_J Jason McDonald

    Thanks for the informative post! I think I’ll try it out on my site as well!

    • http://webcontrolroom.com/ Dan Norris

      Cool Jason hope it helps you out.

  • Mark van den Berg

    I was wondering, doesn’t the second rule in your custom segment example (organic traffic only) render the first rule (major search engine) redundant? Or are you really only interested in natural search traffic from these three sources specifically?

    • http://webcontrolroom.com/ Dan Norris

      Hey Mark. Yes natural search traffic, without the second rule it would also include Adwords.

  • Mark van den Berg

    Your ‘Direct’ segment seems to me to require a statement logic of something like this: source = direct OR (source = organic AND keyword = brand name). But the custom segment tool does not allow for (x OR (y AND z)). It only allows (x AND (y OR z)). Have you found a solution for this?
    thanks!

    • http://webcontrolroom.com/ Dan Norris

      Well I guess there are a few ways to look at it. First off I don’t use Adwords anymore so I’m just using ‘Keywords’ I’m not specifying that it has to be organic.

      If I did use Adwords I probably wouldn’t be bidding on brand keywords anyway so just putting keywords for the second part of the equation is fine.

      If you were using Adwords and you were bidding on brand keywords then it becomes a bit murky because I would still almost consider it direct traffic if someone goes into Google and puts in ‘Informly’ and clicks on one of my ads. It’s not a new customer, it’s someone who already knew about me so I still think it works. But I can also understand the other view.

      I’ve attached an image of what mine looks like.

    • marnusw

      You can swap between AND and OR if you invert all inputs and the output, e.g.
      X AND Y === !(!X OR !Y)
      If it becomes more elaborate you might have to document your changes somewhere to keep track of what you’re doing though. ;-)

      • http://webcontrolroom.com/ Dan Norris

        Thanks for the tip!

  • Guest

    Thanks Dan this is a great resource I’ve never seen this used.

    • http://webcontrolroom.com/ Dan Norris

      Thanks!

  • http://benhebert.com/ Ben Hebert

    Dude Dan this is an incredible resource. Surprised that you’re not getting more love on it! Thank you for taking the time and energy to put this together.

    • http://inform.ly/ Dan Norris

      Thanks Ben the social stats aren’t accurate I first published it on my old domain. It’s one of my most popular posts.

  • Dave Newgass

    Thanks for this Dan! I am always looking for great information about GA from a trusted source. Keep ‘em coming :)

    • http://inform.ly/ Dan Norris

      Good to hear Dave! Thanks for the comment I’m glad you found it useful.

  • http://www.matthewnewton.net/ Matthew Newton

    Hey Dan.

    I’m implementing this right now for the Web Agency Podcast,

    I’m kind of scared/intimidated because honestly.. I’ve never done any conversion tracking besides the most basic level.

    I’ll let you know how it goes

    Mat.

    • http://wpcurve.com/ Dan Norris

      Thanks man it’s easy. The interface has changed now but the concept is the same.

      • http://www.matthewnewton.net/ Matthew Newton

        Segments are easy, the Goals thing.. should be easy but not working. Bashing my head against the desk at the moment.

        • http://wpcurve.com/ Dan Norris

          The URL based goals are much easier. I’m using them now just so i know I can do it all myself and it’s much simpler to understand. I re-direct to a page after signup and that page is the goal in Analytics and it also serves the purpose of requesting more information from the user so we can further segment them in Infusionsoft.

          • http://www.matthewnewton.net/ Matthew Newton

            Hell yeh! I was wanting to use Goals and then I realised I could just redirect all the different signup forms to different thankyou URLs and make it work that way.

          • http://wpcurve.com/ Dan Norris

            Yeah, still use goals though just use a URL based goal. Then you get all the nice conversion reports.

          • http://www.matthewnewton.net/ Matthew Newton

            Cool yeh that’s what I meant. Scuse my misuse of the terms.

            URL based goals are still possible to track all different forms! woo