Coast Digital Blog

Google Analytics makes our lives very easy when it comes to dividing our traffic sources into easily manageable chunks. It automatically splits out non-paid search traffic and paid traffic from Adwords. It also tells us about keywords from all the search engines, plus referrals and direct traffic.

But what happens if you send out an email with links to your site, or run a PPC (pay per click) campaign on Yahoo or Bing? I’ll tell you what: it makes a big mess of all your data.

If you send an email, Analytics will track any clicks either a referral (from web-based clients like Hotmail) or as direct traffic (from Outlook or another offline email client). Your data will be totally skewed.

It gets worse. If you decide to run a PPC campaign anywhere other than on Google, Analytics will record the traffic as natural visits, and not paid visits. This is a real pain, and it ruins your data for natural traffic.

So what’s the solution?

As you know by now, the people at Google are very kind to us. They have provided us with a tool that can generate a tag to stick on the end of any URLs or links we might need to track. If you’re going to track lots of additional marketing activity, Google’s URL builder will become a very dear friend.

This is what it looks like.

And you can find it here: https://www.google.com/support/analytics/bin/answer.py?hl=en&answer=55578

All you need to do is follow the on-page instructions, add the details of where you’re sending the traffic, and then fill in (at least) the source, medium and campaign names. Then hit ‘generate’.

Google will now give you a URL. Any traffic that visits your site via this URL will be automatically assigned to the traffic source, medium and campaign name that you’ve specified.

Example 1 – email marketing.

This is the most common reason I use the URL builder tool. Whenever we send out marketing emails for our clients, we tag all of the links in the email back to the client’s site. This allows us to define the email as a separate traffic source in Analytics. Like so:

Using the URL builder, we put a name for the email in the ‘Campaign Source’ box. You should give your email an easily recognisable name, so when you come to report on it later, there should be no confusion. With email campaigns, it’s sensible to add the send date, as above. 

The ‘Campaign Medium’ should be defined as email. The only other required box is ‘Campaign Name’. If the email you’re tagging is part of a campaign, then you’ll want to use that campaign name. If it’s a one off, we tend to use the source name.

Example 2 – tracking a non-Google PPC campaign

Let’s say you want to track a Yahoo Search Marketing (PPC) campaign in Google Analytics. If you run the campaign without tagging it, Google will automatically record the visits as coming from Yahoo, but it will wrongly assume they are organic clicks. 

To solve this problem, you need to tag your campaigns. Using the Google URL builder, you can specify the source as Yahoo and the medium as CPC (cost per click). You can then specify the campaign as “YSM-PPC-campaign” (or something similar). The URL builder will generate a tag that looks like this:

http://www.coastdigital.com/testpage.cfm?utm_source=yahoo&utm_medium=cpc&utm_campaign=YSM-PPC-campaign

All you have to do is pop these tags on every ad URL in the campaign:

?utm_source=yahoo&utm_medium=cpc&utm_campaign=YSM-PPC-campaign

In this way Google won’t automatically define your CPC traffic as organic.

To gather even more detail, though, you should make use of the ‘Term’ box in the URL builder. This lets you to give every keyword its own tag, allowing you use Google Analytics to measure the success of every single one.

Top Tip: Tagging multiple keywords (some Excel knowledge needed)

Tagging every keyword in your Yahoo account might seem like a daunting task. However, using Excel, you can quickly and easily tag an account of any size. Export all of your campaigns into Excel, and sort the whole spreadsheet by component type. It will then group all of your keywords together. 

Next, use the URL Builder to create a tag that looks something like this (using this tag will work for any Google Analytics user, so feel free to copy and paste). You’ll see the utm_term and utm_campaign are blank.

Scroll across the header row of your downloaded Yahoo campaigns until you find “Keyword Custom URL” (normally column H). Insert a few blank columns next to this row (either side is fine).

In one blank column put the destination URL of the keyword. In the next column add the section of the tag that ends with utm__term=

In the column after that, put the last part of your tag:

Drag these tag columns all the way down so that every keyword has these empty tags in the column next to them. Now go back to your 'Keyword Custom URL' column and use Excel’s concatenate function to tie your columns together. You should tie in your destination URL, then the first part of the tag, then the keyword, then the second part of your tag, then the campaign column. (I’ve colour coded each part so you can see how the formula builds the tagged URL).

=concatenate(destination-url, first-tag, keyword-column, second-tag, campaign-column)

This should give you a fully tagged keyword level destination URL. It will look something like this:

Drag the concatenate formula down your list of keywords and it should automatically fill in all of your destination URLs.

Highlight the ‘Keyword Custom URL’ column, copy it, and then with the column still highlighted press ‘Paste special’ and choose ‘values’. This will overwrite your formula with your new, tagged URLs. 

After all that, you just have to delete the columns you inserted, and upload the changes. 

What else should I tag?

Everything that isn’t already sifted automatically by Google. This will help keep your Analytics accounts as tidy as possible, and allow you to access more information on the performance of your online activity, including elements such as banners, advertorial, or social media.

When you’ve got your Analytics account up and running, you can start tracking the more advanced features on your site. 

Event tracking is a more recent (June 2009) addition to the Google Analytics toolset, and it allows you to track almost anything. 

If your site uses lots of Flash, then event tracking is definitely for you. For example, if you have a Flash music player, you can use event tracking to count the number of times users press ‘play’, ‘pause’, ‘fast forward’ and so on, giving you a great insight into how people interact with your content. Event tracking is also useful if you want to track other user interactions – such as out-clicks or ‘mailto’ links – without affecting or skewing page view numbers.

The code

The code for tracking events is very simple indeed. All you need to do is tag each element you want to track with the following line of code (this example assumes that ‘pageTracker’ is the name you’ve given to the tracking object).

pageTracker._trackEvent('category', 'action', 'label', value);

This code needs to be called in the page after the pageTracker is initialised. Sometimes the best (and certainly the easiest) thing to do, is to move the standard GA tracking code to the top of code – i.e. straight after the opening <body> tag. 

The four fields within the brackets allow you to categorise each action performed on the site. Below is a diagram showing how events are structured when you come to report on them. 

Here’s a bit more information on each of the parameters that you can pass into the code. 

Category. This field is mandatory and should be passed with every event that you track. The category is the highest level in the event tracking ‘architecture’, so it should be used for grouping your events at the widest level. For example, you might have ‘videos’ as the category. This will group all video-related events. Alternatively, you could have ‘downloads’ as the category and group together all download events  

Action. Action is the name you give to the specific way in which the user will be interacting with the site. This field is also mandatory. For example, if you’re tracking video actions, you might name them ‘play’ and ‘stop’. Or, if you’re tracking downloads, then you might use the type of file, ‘pdf’, ‘exe’ or even a specific filename. 

Label. Label is an optional field that can be used to give more information on each action. So, to return to the video example, you could use the video name as the label. You would then not only be able to record all video interaction at a category level and how many video plays there have been at action level, but also how many plays there have been on a video-by-video basis at label level.

Value. Value is also optional. The value needs to be an integer (i.e. a whole number, meaning - unfortunately - that you can’t use cost or revenue numbers). 

Top Tip – If you are using values pay close attention to the tag - pageTracker._trackEvent('category', 'action', 'label', value); You won’t need inverted commas around the value as it’s an integer and not a string. Putting inverted commas around your value parameter will prevent the tag from working.

Using our video example once more, Google suggests that download time could be used here. But if you’re really clever, a better use would be to record how many seconds of the video the user watched before they stopped it. If you get low numbers, you’ll know the video isn’t connecting with your audience. 

Plan Ahead 

It’s really important not to rush into an event tracking implementation. You need to plan exactly what you want to track and then work out how you’re going to structure the events. Getting the structure wrong will mean you won’t be able to get the most out of your stats. 

You also need to think about the future. Ask yourself what else you are likely to be tracking as your site develops. You need to be able to allow for new categories/actions/labels to be added without them affecting your current Google Analytics stats. 

This week Facebook started rolling out changes to the way in which it manages privacy for over 350 million users. This is no small task, and Facebook has done well to keep everything simple and straightforward. When you log in, you are invited to alter your settings. All you have to do is click ‘next’, review your settings and then click ‘finish’. Nice and easy. 

The only problem is with the options. I can either have my OLD settings or give access to EVERYONE: I can’t make my settings more private than they already are. Most of my OLD settings were set to EVERYONE, meaning my options were to give access to certain information to EVERYONE – or EVERYONE! 

So I decided to make my profile a little more private. This is how I did it.

The best way to keep your personal data safe would be to close your Facebook account, although that would be a bit drastic. Instead, you should access the privacy settings under Settings >> Privacy Settings, where you will see the menu below. 

Let’s start with the Profile information section. After clicking this option, you will need to click ‘change settings’ and enter your password. Once you have done this, all the boxes will become active. 

If you change all the settings to ‘Only Friends’, this will ensure your information is not publicly available – only your Facebook friends can see your information. Of course, it’s crucial that you are careful about the friends you make – there are thousands of Facebook accounts set up with the purpose of stealing personal information, and some of your friends may not be so selective. So don’t give access to ‘Friends of Friends’ and certainly don’t give it to ‘Everyone’. It is also important that you set all your albums to ‘Only Friends’.

After all, do you want total strangers looking at your holiday snaps or family photos? Probably not.

Contact Information

Who do you want to be able to contact you? Make sure that you have all these settings set to ‘Only Friends’, but allow other people to send you messages via Facebook and make friend requests. Unless you want to be a complete hermit. In which case, you might want to ask yourself what you’re doing on Facebook in the first place. 

Applications and Websites

This is the most shocking of all the privacy sections. You need to remember that anyone can write a Facebook application, and it doesn’t take much imagination to work out that anyone can write a sinister data collection tool and hide it inside a nice, friendly game. Companies might even collect your data and store it for totally innocent purposes. But they would still be holding your data, and if anyone hacks into their servers or steals their equipment, your data will be compromised. 

My advice here is simple. DO NOT INSTALL FACEBOOK APPLICATIONS. This way your information will not be stolen. 

There’s another risk, though – check out this tab:

What your friends can share about you through applications and websites

If your friend installs Facebook applications, they can also harvest all your personal details. This brings me back to my point about being very careful who you befriend on Facebook. Your friend ‘Dave1234’ might be installing every application available and not realise (or care) about privacy. 

Make sure that you uncheck every box:

Search

Do you want your profile to appear in search engine results? Do you want to be found by any person on the internet? 

Imagine you are looking for a new job and your potential new boss does a quick Google search for ‘Oliver Bloggs’.  He finds your Facebook profile and checks out your drunken holiday pictures and reads the banter going on on your wall and in your status updates. 

This is a serious issue. Privacy affects everyone.  Make sure that you change these settings and stop yourself being found in the search engines. 

Events

When you are creating an event – such as a charity fundraiser or club meeting – remember these can also be viewed and harvested. It makes sense not to disclose personal information here either.

Conclusion and final advice

Facebook does not look after your data by default. You have to actively make the relevant changes to your settings. Update them. Weed out your friends list – you don’t really have 238,024 of them in real life. And always remember that your account could be compromised or taken over by a hacker. 

Three last tips:

• Do not put your telephone number/ address on Facebook. Unless you would like strange phone calls and random internet people turning up at your door.
• Do not make friends with random strangers.
• Do not install Applications that want access to your personal data. Can you ever be sure it’s safe?

The author of this post prefers to blog anonymously – to protect their personal data.

It’s no secret in the Coast Digital offices that I am not a fan of Apple’s iPhone; and that’s why I opted for an HTC Hero on a tariff from Orange. Without doubt, it is the most amazing phone I have ever owned. 

Although my phone is certainly better than your iPhone (if you have one), that’s not what this post is about. Instead, I want to tell you about what happened when my HTC Hero failed.

Last Saturday, I woke up and found my phone had no signal. I tried to make a phone call, but I was told my handset wasn’t registered on the network. I wondered whether Orange was having problems, so I put the SIM card in one of my older phones. It worked perfectly. The problem had to be with the HTC Hero.

At first I tried changing settings, and then decided on a factory reset – I reasoned that I might have installed a dodgy application or something. No joy.

So I called Orange. 

A very helpful person offered to replace my handset, and gave me a range of times I could receive my replacement on the next day – Sunday.

I don’t have Orange Care. I am not a business user. And when I chose for my replacement phone to be delivered between 9am and 1pm, it arrived at 11am sharp. All I had to do was plug in my SIM card and – bingo! - I was back on the Orange network. After re-syncing my contacts with Outlook, I was good to go.

From broken phone to replacement – in only 19 hours. Over a weekend. That’s customer service.

It’s also good business: so many of our clients come via referrals from satisfied customers. Perhaps we can help you too?

What are goals?

Google Analytics ‘goals’ allow you to measure specific actions that users carry out on your site. They can track events like a submitted enquiry form, a brochure download or a completed transaction. That’s why it’s important to set up sensible and comprehensive goals from the outset – only then will conversion rates be calculated and shown in Analytics reports.

Goals can be triggered by one, or a combination of activities that include page views, button presses and form submissions.

You can use up to 20 goals, grouped into four sets of five, within each Analytics profile – if you need to track more then you’ll need to add an additional profile to your account. But don’t worry: 20 is more than enough for most purposes.

Setting up page view goals

Goals are most commonly triggered by a page view. For example, a ‘thank you page’ view will indicate that a user has completed your contact form and been directed to the next stage.

So, how do you measure something like this in Analytics?

First of all, navigate to your list of website profiles and select ‘Edit’ next to the name of the website you are working on.

You’ll be taken to a screen on which all your available goals are listed.

Click on ‘Add goal’ and then give your goal a clear name – the easier it is to recognise in your reports, the better.

Next, select ‘URL Destination’ as the goal type. We will go on to look at the other options at the end of this post.

Match Types

At this point you will be asked to enter a ‘match type’. You have three to choose from.

Head Match will match your goal URL to the leftmost part of the page URL – this means that dynamic addresses can be tracked. For example, where a unique identifier is used in the URL /thankyou.html?id=1234567 a goal would be triggered if the Match Type is set to ‘Head Match’ and the goal URL is set as /thankyou.html

Exact Match matches the goal URL only to the exact page URL that appears in your reports.

Regular Expression Match allows URLs that use the same regular expression to trigger the same goal. For example, you could ensure that dynamic URLs such as www.mysite.co.uk/?editpage=11567830 and www.mysite.co.uk/instock/?editpage=90385382 both triggered the same goal – both use the regular expression ‘editpage’.

Top Tip: if you need to test which goal URL to use, either because you have dynamic URLs or a number of pages that relate to the same goal, you can construct a RegEx expression in the Top Content report filter box (at the bottom of the Top Content Report page). This will show the page views that will trigger a goal using that expression. Simply copy and paste the expression into the Goal URL section and select “Regular Expression Match” as the match type.

Goal Values

It is possible to assign a monetary value to a goal. If a completed goal has a commercial outcome, it can be assigned to the goal value. If you know you have an average online order value of £500, and you know that about 10% of enquiries result in a sale, you might assign £50 to your goal value.

It is worth noting that, whilst goal values give you an indication of how much your traffic is worth, the function is not as versatile as e-commerce tracking, which is covered later in this blog series. E-commerce tracking can attribute values and specific products to transactions, but there’s nothing to stop you using Goal Values as well.

Using additional steps and funnel visualisations

By including additional steps in the goal process, it is possible to monitor the whole process in a visual format. For example, you can gather statistics for every stage of a purchase – the number of people who add items to a basket, the number who go on to the checkout and the number who complete their purchases. You can add up to ten steps in these ‘funnels’.

In this way, you can see the points at which visitor drop-out is occurring and identify possible weaknesses in the funnel. If you can identify the pages that users are navigating to instead of the ones you want them to go to, it will help you optimise the buying process – and ultimately provide information that can help you to improve conversion rates.

The funnel visualisation:

Engagement based goals

In its most recent update, Google added the option to use different goal types instead of the standard ‘URL Destination’ goal.

You can now measure goals that rely on ‘time on site’ and ‘pages/visit’. These allow you to record goals when a certain threshold has been reached (either greater or less than the predetermined value) for a set time on site, or a number of pages per user visit.

Finally, once your goals are set up and they have collected data, you’ll see new statistics in your visitor, content and goal reports – allowing you to analyse goals within different segments of your Analytics data.

Thanks to goals, you are not only able to measure how effective your website is, but you can use the data you collect to refine the way it works. It’s just one part of what we call ‘marketing you can measure’.

In October Google began experimenting with a new look – a minimalistic approach designed to that give the user what they wanted without distraction. It was dubbed Google Fade, and it works by only displaying the Google search bar, the Google Search button and the I’m Feeling Lucky button until the user moves their mouse. At that point, a reverse fade brings everything into view.

Now you see them.

And now you don't.

 

All in all, we ran approximately 10 variants of the fade-in. Some of the experiments hindered the user experience: for example, the variants of the homepage that hid the search buttons until after the fade performed the worst in terms of user happiness metrics... However, in the end, the variant of the homepage we are launching today was positive or neutral on all key metrics, except one: time to first action. At first, this worried us a bit... Then, we realized: we want users to notice this change... and it does take time to notice something (though in this case, only milliseconds!). Our goal then became to understand whether or not over time the users began to use the homepage even more efficiently than the control group and, sure enough, that was the trend we observed.

 

Since it was launched worldwide on 3rd December, there has been a user backlash regarding the new design. On the official Google help forums, many threads are filled with complaints. 

 

Some threatening:

 

“Leave the Homepage alone, or else I'll find another search engine to use.....”

 

“Worst Google idea ever. I am about to go find a new browser for my home page. They created the perfect solution to a non-existent problem.”

 

Some frustrated:

 

“Seriously. I couldnt even beleive [sic] it myself, I thought I was hallucinating!!! Then finally I arrived here, to find out fellow googlers are also experiencing the same atrocious disaster. It is extremely slowing me down as I am a Sr. Network Engineer and have multiple windows open at all times, and opening new ones all day long. My homepage being Google, is now taking an extra couple seconds to come-up, and when you are working with seconds of time, that can be DEVASTATING.”

 

And some just downright funny:

 

“Stop experimenting with the stupid "fade" already. It isn't better and it isn't necessary. Stop making changes just for the sake of making changes. A search engine isn't supposed to be a video game. The only folks I can imagine thinking this is 'cool' are teenagers and twenty-somethings with the attention span of a housefly.”

 

Personally I don’t see what the fuss is about. It looks nice, and I know how to find the little 'Mail' icon without any fuss. However, I am intrigued to see this fade effect tie in with the upcoming Google design, as this change would then be a much more radical transformation from the Google we know and love.

 

The future?

 

Welcome to the first post in a new Coast Digital series designed to help you unleash the power of Google Analytics (GA), the enterprise-class web statistics service.

Every Tuesday and Friday over the coming weeks, we’ll be covering everything from the absolute basics of GA to some of its more advanced features. We’ll be showing you exactly what data can be tracked with Analytics, and we hope you’ll pick up useful insights that will transform your website’s performance.

Sound good? Then let’s get started.

The First Step

We’ll begin with the first task you’ll face after deciding to use Google Analytics: setting up your account.

To do this, head over to www.google.com/analytics and press ‘Sign Up Now’. If you don’t have a Google account, you will need to create one. If you do, use it to sign in.

When that’s done, you’ll be taken to this page and asked to click on another button to get started. You'll then see this screen.

On this page, you’ll need to enter your web address and give your account a name. The account name is principally for your own reference, so it’s worth picking something that’s recognisable to everyone who will use it. 

It’s also important to add in the correct time zone for your account.

Top Tip: If your site specifically targets multiple time zones, set up the account with your local time zone to keep things clear.

Finalise the account details with your name and location:

And then we get to the exciting part...

The Tracking Code

Google Analytics tracking code comes in many shapes and forms, and is used for different purposes. We’re going to cover the more advanced versions later on in our series.

However, in this post we’re going to cover the absolute basics. All you need to do is select the code in the grey box and copy it. 

When you’ve done that, you’ll need to paste it into the source code of any pages on your site that you want to track. Ideally, the code needs to go in before the closing </body> tag.

Top Tip: If your site uses a template, then putting the code in the header or footer is a good idea. That way, you’ll be able paste the code into one file and the template will automatically add it to every page on the site.

Once you’ve pasted your tracking code into the site, press the ‘Save and Finish’ button. And that’s it! Your Google Analytics account has now been created, and should start showing data within a few hours.

Coming up on Tuesday (8th December): Goals and funnel visualisation.