marketing you can measure

The blogCoast Digital Blog

Unleash the Power of Google Analytics (10): Troubleshooting Analytics

 Analytics Part 10: Troubleshooting Analytics

Sometimes, a Google Analytics installation doesn’t behave quite how you might expect it to, showing inconsistent data or just behaving strangely.

It isn’t always easy to see what might be causing the problem, but at Coast Digital we often come across accounts that have been set up incorrectly, or have crucial tracking code mistakes.

That’s why we’ve decided to share of the more common problems that we encounter, what the symptoms are and how best to fix them.

Problem 1: There’s no data in my Analytics account

More often than not, this is caused by missing, wrong, or incorrect tracking code.

The Google Analytics tracking code is account and profile specific, so make sure that the right code for the right account is installed on the right site.

To minimise potential problems, update the older urchin.js tracking code to the new ga.js code – or, if you’re feeling adventurous, use the upcoming asynchronous tracking code instead.

No data in Analytics

Problem 2: An overall drop of traffic from all sources

If this happens, the tracking code is probably missing from some pages on your site.

Check the traffic levels to specific pages to determine which ones are causing the trouble, and then repair them.

It’s easier to spot offending pages if they are among the most visited, but can be trickier if the tracking code is missing from a variety of less popular pages. It’s also worth checking to see whether your tracking code is missing or incorrectly set on any subdomain – we’ll cover this scenario a bit later on.

Problems with the tracking code can occur very easily if your site has a number of users who have access to your content management system (CMS) - particularly if they are able to a site where a number of users have access to the Content Management System, can revert to old versions of page content easily, or have access to the template code.

Top Tip: there is a very handy Firefox plugin called “GA?” that will show you at a glance whether a page has Analytics tracking code installed on it. Download it from the Firefox add-on site at https://addons.mozilla.org/en-US/firefox/addon/5631

Problem 3: No AdWords cost data is showing

This often occurs when your AdWords account is not linked to your Analytics account.

There is a manual setting that must be applied before cost data is shown.

To check that it is active, log into your AdWords account, click on the ‘My Account’ tab, and then on ‘Account Preferences’. Beneath ‘Tracking’, you will see the ‘Auto Tagging' setting’ – this should be set to ‘Yes’.

Apply Cost Data

You should also check that your AdWords account is linked to the correct Analytics account – if it isn’t, no data will be collected.

Problem 4: No Keyword or Ad Group data for non-Google PPC traffic

Although AdWords can – with a single setting – share cost and campaign information with Analytics, it’s not as simple if you want to do the same with MSN/Bing, Yahoo! or another provider.

If you do want to do this, you need to make sure that each destination URL is tagged correctly, allowing it to pass keyword and campaign information to Analytics. For details on how to do this, see the earlier post in this series, ‘Tracking External Sources & Tagging URLs’.

It is worth noting that cost data from other providers cannot be passed into Analytics.

Problem 5: A drop in traffic from one PPC source

You can usually pinpoint this problem to a tagging issue. Destination URLs may have been changed without the the correct UTM parameters being appended, or there may be a problem with the PPC account – ads may have stopped running for billing reasons, for example.

Problems can easily occur where there is a number of people responsible for maintaining PPC accounts – they may make changes to advert destination URLs and keyword destination URLs, so it is important that you share your tracking techniques with the everyone concerned.

Problem 6: An entry labeled ‘(Other)’ appears in your content report

If your Analytics account is tracking more than 50,000 unique URL or page views, then all pages that don’t appear in the most popular 50,000 are bracketed together in a section called ‘other’.

One way to prevent this happening is to exclude certain dynamic parameters within your Analytics settings, group like pages together, or run filter sets to that different profiles within the account to report on different segments of data.

Problem 7: Spikes in direct traffic that coincide with bulk email sends

If this happens, it’s probably because email traffic is being reported in Analytics as ‘direct’ traffic – it’s a result of untagged links in emails.

Normally, visits from email clients like Outlook - which open a web browser when you click on a link - are registered as direct traffic. In addition, some webmail traffic - notably Hotmail and Gmail - will also be recorded as ‘referral’.

Once again, the solution is to add tags to email URLs - refer to ‘Tracking External Sources & Tagging URLs’ for details on how to do this.

Problem 8: Goals aren’t appearing

If no goals are showing up in the relevant Analytics report, make sure that your goal URLs are correct. All too frequently, we see the full URL path entered, including the http:// - you need to omit the domain name.

For example, if your goal is at the URL http://www.mysite.com/thankyou.html, you should use '/thankyou.html' as your Goal URL.
 

Goal details

You should also check that your ‘thank you’ page resolves to the correct address – if someone changes the location of the ‘thank you’ page, even though conversions may be coming through to your CMS, database or email inbox, they won’t appear in Analytics.

Enquiry dropoff

Problem 9: A significant number of referrals from your own site

A certain number of legitimate self-referrals should be accepted as ‘normal’ in an Analytics account – particularly from visitors that let 30 minutes pass before clicking on another link on site.

However, if self-referrals form a significant segment of your traffic, a number of things could be to blame:

  • Missing or broken tracking code on some pages. If a popular page is missing tracking code, visitors that hit that page can subsequently be tracked as referral visitors from your own domain name. Identify these pages by looking at the content report to identify missing pages and make sure the correct tracking code is installed on them.
  • Multiple domains and cross domain tracking not set up correctly. If your site spans across an external domain – such as a checkout on a third party provider’s site – tracking must be set up to pass data between the two. They both need to have the same Analytics tracking codes on, with _setAllowHash(false) and _setAllowLink(true) lines added to the code. The referral page should also have each link tagged with _link(). See Advanced Setup Process and Filters for more details.
  • Multiple subdomains. If the secure area of your site sits at a subdomain, such as secure.mysite.com, the Analytics tracking code should be updated to include the line _setDomainName('.mysite.com'), to ensure that cross sub-domain tracking is enabled. Once again, see Advanced Setup Process and Filters for help.

 

Subscribe to our blog and follow future series like this one

Google Buzz gets us all waving

Google Buzz

The internet is Buzzing at the moment with the announcement that Google has introduced a new social feature to its ever-popular Gmail service. It’s called Google Buzz, is linked directly to your existing Gmail account and is easily accessible from the main window.

Gmail Buzz

Inbox BuzzBuzz offers the usual range of social networking features, but promises to “go beyond status messages”. I have had a quick play around with it this morning, but as it is being rolled out slowly it has been a bit hard to “buzz” many of my contacts.

Luckily, I managed to rope my colleague Hannah into Buzzing – just to see what it was like. Happily she agreed, and I witnessed a very nice Buzz real-time update that happened without the need for refreshing.

Friends' comments

Having mastered text updates, we tried a text and photo update. Incredibly, this worked as well! It is also extremely easy to use, using either photos from your computer or from your library.

Picture update

Google has decided to encourage privacy by giving you the ability to share your updates publicly or privately. I like this feature a lot – if, for example, I have a quick update I would like to share with only the Coast Digital team, then I can. This appeals to me a lot more than sending out an office-wide email.

Privacy

Another feature that I am looking forward to trying out is the ability to connect Buzz with other sites, allowing you to pull in your Twitter feed or your public Picasa albums. I use Gmail and Twitter every day, so having them both in one place does seem pretty handy.

Other sites

There is a wealth of other things I should talk about such as complete inbox integration, iPhone Buzzing, and picture importing – but I shall save those until I get fully used to all of this and have a few more followers.

My colleague Hannah Rampton says this about Google Buzz: “I think that Buzz is just a more personal, simpler version of Wave. I love the real-time integration with no need for refreshing.”

And here is Google’s official announcement:

 

Coast Digital Blog Unleash the Power of Google Analytics (9): Advanced Segmentation

Google Analytics: Advanced segmentation

Advanced segmentation is a relatively new feature of Google Analytics. It’s very powerful and can help you unearth some really specific information about your site's visitors.

What is an advanced segment?

An advanced segment is a real-time slice of the data that is in your Google Analytics account. Segments, just like filters, allow your GA profiles to contain specific groups of users, based on a set of chosen criteria.

The advantages of advanced segments over filters are that:

  • you can apply (and remove) a segment on the fly
  • they don’t make any permanent changes to your profile’s raw data
  • they work retrospectively (you can apply them to old data).

How do I setup an advanced segment?

Login to your GA account and visit the dashboard for the profile that you’re going to be using.

In the top right-hand corner, just below the orange bar, it says ‘Advanced Segments’. Clicking on the ‘all visits’ button next to this will open the correct menu.

Advanced Segments Setup

You can then choose from the list of existing segments, or click ‘create a new advanced segment’ if the segment you want isn’t already available.

Let’s assume you’re going to create a new segment. When you reach the next screen, you can select from the available dimensions and metrics panels, drag them in to the main window and use them to create the logic that will segment your data.

As you can see in the example below, I’ve created a segment that will contain visits where revenue was greater than £50. To make sure your logic is right, you can hit the ‘test segment’ button. If no visits are returned by the test, it’s likely your logic is wrong.

Creating a segment

Some example segments

Revenue

The example above gives us a good start. Looking at visits by revenue value can give you a good idea of how your biggest spending customers are using the site.

It’s also useful to look at those users that don’t spend any money on your site. Doing so may throw up obvious areas where the site can be improved – you should be looking for any pages that have high bounce or exit rates.

Goals

We find the segment below really useful. It allows us to look specifically at users who started a goal process, and didn’t complete it. This is particularly useful if you are tracking a shopping cart funnel, or a sign-up process.

Goal starts versus completions

Again, looking for high exit rates is vital. Finding where users are hitting problems and fixing any issues can help increase your conversion rates no end.

Traffic sources

Segmenting by traffic source is one of the simplest uses of advanced segments. Google Analytics gives you a selection of these to use as default.

If you’re paying for traffic to your site, it’s important to make sure that every click counts. There is a pre-existing segment for paid search traffic, but if you’re paying for other traffic via banners, emails, or affiliates, it’s worth investigating how these users interact with your site. Here’s an example:

Email traffic

With these paid traffic segments, you should be paying special attention to landing page performance and other areas of content that are heavily visited.

You can take these segments one step further by combining them: let’s look at visits from a source that didn’t generate revenue.

Expanding the above examples, we can say medium is ‘email’, AND revenue is ‘0’:

Medium is email, revenue is zero

The more ANDs that you include in your logic, the smaller your dataset is likely to be. Make sure that you don’t impose so many conditions that you end up making big decisions on tiny data sets.

Experimentation is the key

As with a lot of Google Analytics functionality, the best way to learn is to gets your hands dirty. Experimenting with segments will allow you to see what options are available and how you can imaginatively manipulate your core data to get the stats that you want!

Subscribe to our blog and follow this Analytics series

The opinions expressed herein are the personal opinion of the author and are not intended as statements of fact and do not represent the view of Coastdigital Limited in any way

RSS FeedRSS Feed