Although the correct tagging of your campaigns seems something basic, I still see a lot of companies and websites that aren’t tagging at all or do it in a wrong way. In this article, I will explain the major tagging issues and how to tag correctly.

Campaign Tagging Problems

The most important issue with tagging your campaigns is the involvement of many different colleagues and partners.

With the implementation of a Google Analytics code or the configuration of goals, you are most of the time depending on one or two people.

With the tagging of your campaigns, often multiple people are involved because the search marketer, affiliate marketer, email marketer, display marketer, social marketer, external banner, display partners, etc., all have to tag their campaigns. More people involved means a bigger chance for errors. Even if all of your colleagues are tagging, it is unlikely that they all use the same notation.

For example, email marketing can be described as e-mail, email, Email, E-mail or mail. If not everyone is using the same notation, Google Analytics will report multiple notations for the medium email marketing. This makes a correct way of analyzing the data more difficult.

The use of capitals can be solved by configuring lowercase filters on the different parameters. These parameters will make sure that the mediums email and Email will return in Google Analytics as email.

More on the use of lowercase filters can be found here: http://www.google.com/support/analytics/bin/answer.py?answer=55593.

To solve the different notations of the word email, a tagging plan is required. (Although this could also be done by the search and replace filter, I would recommend using a clear tagging plan.)

Create A Tagging Plan

Most of the people who read Search Engine Land understand what campaign tagging is and why it is important; however, some people who have to use campaign tagging don’t. This is exactly why it is necessary to start with an explanation of what campaign tagging is and why it is important.

The next step is to identify all colleagues involved in the campaign tagging process and to make a clear structure of how to tag campaigns.

Before starting with an example, it is important to know that there are five campaign tracking parameters of which three are required (utm_medium, utm_source and utm_campaign):

• utm_medium: identifies the channel of a campaign (i.e. email, affiliate, display)
• utm_source: identifies the source within a channel (i.e. name of an affiliate network)
• utm_campaign: identifies a campaign within one or more sources / channels (i.e. Newsletter April)
• utm_content: extra parameter for some extra information regarding content (i.e. banner1, textlink2, bottom link)
• utm_term: is specific for non-AdWords paid search and gives you the possibility to identify the keyword

In the following example, the traffic comes from email marketing, affiliate marketing, display marketing, social media and external partnerships. I will only use the required parameters, but for even deeper analysis you could also make use of the other parameter(s).

Email Marketing

For email marketing, you might use the following tag for your campaigns:

?utm_source=newsletter&utm_medium=email&utm_campaign=newsletter_week20

Within Google Analytics, you can then find the collected data for this under All Traffic Source as “newsletter / email”.

Normally the Email Marketing Software that you use can place the tag that you want automatically after all URLs you use for your email marketing campaigns.

Variable parameters

• utm_campaign is variable

For each newsletter that you send, change the week or month number, so you can identify which newsletter it is. When you analyze the data, click to Campaigns and you can see the results of the different newsletters.

Affiliate Marketing

For affiliate marketing, use the following tag:

?utm_source=commisstion-junction&utm_medium=affiliate&utm_campaign={affiliateID}

Within Google Analytics, you can then find the collected data under All Traffic Source as “commission-junction / affiliate”

Normally, the Affiliate Network you work with can place the tag you want automatically after all Affiliate URLs. They also can make sure the affiliateID is tagged.

Variable parameters

• utm_source is variable — fill in the name of the Affiliate network you work with.
• utm_campaign is variable — this will be the affiliateID.

Some examples:

• utm_source=commission-junction
• utm_source=clickbank

Partnerships

For partnerships, you might use the following tag:

?utm_source=startpagina&utm_medium=partnerships&utm_campaign=springsale

Within Google Analytics, you can then find the collected data under All Traffic Source as “startpagina / partnerships”.

If you make special deals with a company to promote your product, you of course want to know the results of your investments. You can tag these partners and analyze the results.

Variable parameters: 

• utm_source is variable. Fill in the name of the partner you work with.
• utm_campaign is variable. Fill in a specific product promotion or strategic campaign.

Display Marketing

For Display Marketing, you might use the following tag:

?utm_source=sanoma&utm_medium=display&utm_campaign=springsale

This way in Google Analytics, you can then find the collected data under All Traffic Source as “sanoma / display”.

If you make a deal with a Display Network to promote your product, you of course want to know the result of your investment. You can tag these Display Networks and analyze the results.

Variable parameters

• utm_source is variable. Fill in the name of the Display Network you work with.
• utm_campaign is variable. Fill in a specific product promotion or strategic campaign.

Social Media

For Social Media, you might use the following tag:

?utm_source=twitter&utm_medium=social&utm_campaign=tweetoffercanada

Within Google Analytics, you can then find the collected data All Traffic Source as follows “twitter / social”.

You can track the tweets you send out by adding a tag like above to your URL’s on Twitter. It is also possible for other social media such as Facebook.

Variable parameters

• utm_source is variable. Fill in the name of the Social Network.
• utm_campaign is variable. Fill in the name of the tweet.

Please note that this is just a guideline and that you may have to adjust it to get the best results for your marketing campaigns.

When you have finished the tagging plan, send and explain it to everyone involved and keep checking the data. Don’t forget to mention the Google URL Builder for an easy way to create the URL’s.

Please let me know what your suggestions are regarding the tagging of campaigns.

Of course, this ideal situation doesn’t happen at a lot of organizations. They are working with half-year or yearly SEA budgets set at the beginning of each year.

These search advertising budgets are mostly divided by the number of months and maybe a little extra in the busiest months of the year. But is this really the best way to divide a fixed SEA budget in order to get the highest ROI?

In this article, I will explain how Google Analytics can help you in determining how to divide your SEA budget across the different months in a year.

Trend-Sensitive Markets

Most organizations operate in a trend-sensitive market, although some in low trend-sensitive markets (like online supermarkets) and some in high trend-sensitive markets (like online travel). Related to these trends are the search volumes and the number of clicks an organization can get from Google, Bing or any other search engine in a specific month.

If the SEA budget is evenly divided across the months, it could happen that in one month, the budget is depleted on the 15th and in another month, there is still budget left at the end. This clearly isn’t the way to get the highest possible ROI.

Budgeting Using Google Analytics

In order to use Google Analytics for this example calculation, we need a full year of accurate data. In the figure below, we see the visits data for a trend-sensitive market like travel, with peaks in January and around July.

For the example calculation, we will use a yearly SEA budget of $600,000, which would normally be $50,000 a month.

In this case, we will use the export function within Google Analytics to get the monthly figures.

After this, we will calculate the new budget for each month with the following equation:

(number of visits for a specific month / total number of visits in 1 year) * total SEA budget = adjusted SEA budget for a specific month

If we execute this calculation on the monthly visits from Google Analytics we will get the following table:

As we see, the numbers show a huge difference between a normal distribution of the SEA budget and a distribution based on Google Analytics data. With the calculation based on Google Analytics data, we are better prepared for the trends in the market and don’t run out of budget in the middle of a month.

A small note to this calculation is that we are using “old” data and it is only a prediction of what the trend could be in the future.

You can make the same calculation with Google Insights for Search data. If we look at the graph for the US in 2010 on the keyword “sunglasses” we see the following figure:

The sunglasses industry is obviously also a trend-sensitive market. What we could do is download the data from Google Insights for Search as a CSV. This data on a weekly level shows the index numbers for a specific week.

If we want to calculate the SEA budget for a specific week, we  could use the following equation:

(indexnumber for a specific week / total sum of indexnumbers for all weeks) * Total SEA budget = adjusted SEA budget for a specific week

I think this article may help you to better divide your search engine advertising budget across the year and to get even higher ROI on you SEA campaigns. Please let me know in the comments how you calculate the distribution of your SEA budget across all months.

A few of the companies I work for never looked at these reports or didn’t think they were interesting enough to look at. To make the reports more interesting for them, I created an Excel sheet in which I analyze which revenue potential they are currently missing. So which Google Analytics reports am I talking about? Here they are:

Java Support

The first report I want to cover is the Java Support report which can be found within the Visitors > Technology part of the new interface of Google Analytics. This report shows what percentage of your visits had Java Support and what percentage of your visits didn’t.

The following figure shows that 88.92% of the visits in this particular case had Java Support and 11.08% of the visits didn’t had Java Support. This is not something to get excited about, but it gets more interesting when you dive deeper into the data and select the ecommerce conversion rate for the two groups of visits.

What you can see in the second figure is that the e-commerce conversion rate for visits with Java Support is 1.48% and without Java Support, this is only 0.24%. This is a huge difference!

You may now have the attention of your manager or client, but to convince them even further, we are going to calculate what your manager or client is currently missing in revenue dollars due to the fact that visits without Java Support convert that badly.

The information I use for this case included the following stats from the selected time period:

What we are going to do is to calculate what the maximum extra revenue potential could be when visits without Java Support convert as well as visits with Java Support. We can serve a different landing page to people without Java Support for example, or we can adapt the page in such a way it works fine for visits without Java Support.

We will calculate this with the following equation:

(current number of visits * conversion ratio visits with java support * average order value) – (current number of conversions * average order value) = extra revenue potential

(779,521 * 1.48% * $600) – (10,484 * $600) = $6,922,146.48 – $6,290,400.00 = $631,746.48

This means an extra $631,746.48 in potential revenue when we can adapt the website in such a way that visits without Java Support convert as well as visits with Java Support. It is just a basic report and a basic analysis, but his should be a number your manager or client has to listen to and take action upon.

Browser Report

We can do the same trick for the browser report.

In the following example, I checked the largest four browsers with a significant amount of traffic. To get these browsers in one nice overview, I selected the advanced filter option (highlighted in the figure below) and for this case I filtered on all browsers with more than 20.000 visits in the selected time period.

The result is the following figure, which shows the e-commerce conversion rates for Internet Explorer, Firefox, Chrome and Safari. All browsers are responsible for at least 10% of the total visits and do represent a significant amount of traffic.

When we use the same formula as we did for the Java Support report, we will get the following equation:

(current number of visits * conversion rate of the highest converting browser * average order value) – (current number of conversions * average order value) = extra potential revenue

(779,521 * 1.41% * $600) – (10,484 * $600) = $6,594,747.66 – $6,290,400.00 = $304,347.66

This means an extra $304,347.66 in potential extra revenue when we can adapt the website in such a way that all browsers are converting as well as the highest converting browser.

Of course, this example does not take into account that some browsers tend to be used by totally different audiences with different needs and wants. It might therefore be unrealistic to get the full extra revenue potential, but such an analysis shows the potential value some organizations are currently missing.

Besides focusing on creating a better checkout process, these reports show it is also very interesting to check whether your website shows great differences between browsers or visits with or without Java Support.

You can do the same for screen resolution, etc, and you can use the same type of analyzing I just did for Java Support and Browsers. Please let me know if you are already optimizing your website based on these reports /analysis or if you have tips of how you are using these reports to convince your manager or client.

• how to create (multiple) dashboards
• the possibilities of the new dashboards
• a few examples for your own dashboards
• suggestions for new features

Multiple Dashboards

The new Google Analytics gives you the power to create up to 20 custom dashboards for each profile. This makes it perfect to create a separate dashboard for each department in your organization or for each traffic source.

The figure at the right shows an example of one of my own accounts with seven different custom dashboards; one for the total overview and six dedicated to a specific traffic source.

This gives the user the possibility to get a rapid and good overview of what each traffic source is doing in terms of the selected metrics like visits, bounce rate, goals, revenue, etc.

For a deeper analysis of a specific traffic source, the user can dive deeper into the different Google Analytics reports or create custom reports.

Adding Widgets

The four different widgets gives the user  a lot of flexibility in creating your own personal dashboards:

• Metric. Shows the value for a single metric, a spark line and what the percentage of the selected metric is compared to the total over a selected period of time.
• Pie chart. Shows a breakdown for a single metric grouped by a dimension with a max of six slices.
• Timeline. Shows a graph for one or two metrics over the selected time period.
• Table. Shows a table with one dimension and up to two metrics over the selected time period with a max of 10 rows.

If you combine these different widgets with the filter options available for each widget, the options are great in creating your own customized dashboard.

Dashboard Examples

Before creating your own dashboard, it is important to clearly define your KPI’s and metrics for each dashboard you are going to create. If KPI’s aren’t clearly defined, you will end up with the wrong metrics or with too many metrics.

The figures below show the different metrics for one of the SEA (Search Engine Advertising) dashboards I created. It shows the overall CPC visits and revenue, transactions divided by a custom variable, most successful campaigns based on visits, visitors, transactions, revenue, as well as the keyword performance (branded vs non-branded) based on visits and bounce rate.

This dashboard gives you a complete overview of the best campaigns/keywords for the selected time period. If you mark the option “compare to”, you will also see the results for the same period before, which allows you to analyze which campaigns / keywords are doing better or worse.

As I mentioned before, the filter options are great, specifically the regular expressions give you a lot of power. They make it possible to make a specific widget for almost any piece of data you want.

The widget below shows a report with the transactions and revenue for non-branded CPC keywords. I did the same for branded CPC keywords and also with other metrics, such as visits and bounce rate. The last option, “Link to Report or URL” gives you the possibility to link directly to a standard Google Analytics report or a URL within the widget.

Here are a couple of  suggestions to create a customized dashboard for SEO:

• Visits, revenue, transactions for organic keywords (branded & non-branded)
• Relevant metrics for top landing pages
• Total for organic visits, revenue and transactions for the selected time period compared to the previous time period

A clear and correct tagging policy is necessary to get the most out of e-mail, affiliate and social dashboards.

Once this is done and you have created all the dashboards, it is really easy to do a quick analysis on every traffic source. For even further analysis, just dive deeper into the data by clicking the linked report / URL or another report within Google Analytics.

Feature Suggestions

Due to the beta release, some options from the old interface are not yet available. At this stage, it isn’t possible to export or share your dashboard, which is really annoying if you want to create the same dashboard for multiple logins. You have to do every widget again and again.

Quickly adding a report to your dashboard is also not (yet) possible, everything is now done by widgets. These widgets are currently limited in regards to mobile options. At this point, it is only possible to see statistics for all of your mobile traffic, while for deeper analysis, it could be interesting to see your revenue or conversion rate for specific mobile devices or carriers.

I think this post should give you same basic insights of the possibilities with the new Google Analytics dashboards. Please let me know in the comments below how you’ve been using Google Analytics dashboards, what features you feel are missing and what you’d like to know more about.

Avinash Kaushik

Avinash Kaushik is the Analytics Evangelist for Google, co-founder of Market Motive and author of inspiring books like Web Analytics 2.0 and Web Analytics: An Hour A Day. Besides that, Avinash is a well known speaker at the largest online marketing conferences around the globe and his own blog Occam’s Razor is a must read for every marketeer / web analyst.

What do you think were the three most important developments in web analytics in 2010?

I am a little biased here but one of the ones was the introduction of intelligent features into tools like Google Analytics. If you look at Analytics Intelligence and Weighted Sort they both make the job of an analyst incredibly easier by proactively applying math and advanced algorithms. Our tools have been too dumb for too long, I am glad it is changing.

Secondly, I am super excited about the massive explosion in affordable qualitative analysis tools like Kiss Insights, UserTesting.com and ConceptFeedback.com. Finally we can get voice of customer into our daily decision making in innovative ways that were not really possible before.

Finally I love the entire social media data explosion. Social media is a lot less exciting than we think (and I say that as someone who has over 30,000 followers on Twitter!), it is also distracting many marketers for strategies that they should be paying attention to. But there is no doubt that social media throws a big wrench in the way we have measured success, or how we have defined success. I love proliferation of tools and attempts to figure out how to measure what matters (I attempted to in my blog post Social Media Analytics.

What are the most important topics in 2011 for web analytics?

Here is what I think it will be: We will continue to obsess about technical implementations, complain about how no one takes action on any data we provide and be distracted by throwing up massive reports on any data that can be gotten out of Google Analytics or Omniture or NedStat.

Here is what I hope happens: We get out of the silly data puking business and realize that our purpose in life is to solve business problems, then we go on to embrace business centric structured approaches like using the Web Analytics Measurement Model. We, finally (!), embrace testing and optimization as a way of life. Yes it is not easy to deal with, but then again generating millions of Euros of improvement in our businesses never is. We will see some analytics tools disappear (and I mean in the broad Web Analytics 2.0 sense), others appear and others still become better.

What will be the influence of the “do not track” options now being made available in web browsers?

Privacy is a very important concern. Unless users have confidence that websites have been transparent about what data is being collected and how it will be used, it will be hard to stay in business. The development of strong privacy options in browsers, strong privacy opt-out plugins by companies like Google and other actions like that increase confidence of our users. This is a good thing. On the web we are blessed to already collect more permission based anonymous data than we know what to do with. All of us analysts, marketers and website owners will continue to have data to have sufficient permission based anonymous data to improve the ROI of our online programs and investments.

Jim Sterne

Jim Sterne is the author of several marketing books, his most recent, Social Media Metrics: How to Measure and Optimize Your Marketing Investment. Besides that, Jim is the founder of the eMetrics Marketing Optimization Summit and co-founder of the Web Analytics Association.

What are the most important topics in 2011 for social web analytics?

I make a distinction between social media metrics and web analytics. Web analytics is all about measuring the behavior of people on your website. Social media metrics looks at measuring the activities out in the “sociosphere.”

In short, the most important topics will be:

• How do I quantify publicly expressed sentiment and use that information to make business decisions?
• How do I determine if my social media activities are producing value?
• How do I integrate data from social media and on-site behavior and offline customer data to optimize my marketing efforts?

This last one is the most important. Marketing data integration was our main goal in the past few years until we got sidetracked by social media.

More and more internet activity will move from traditional websites to social media sites, apps, etc. What will be the impact of this on the traditional collection of data as we know it?

A healthy web analytics industry has developed complex tools over the past 15 years. These tools allow organizations to instrument their own websites at will and glean a considerable amount of information from them. Now that people are building useful applications (including commerce) and entire websites inside social networks, they will become reliant on those networks for that data. The level of data granularity will diminish for a few years until larger organizations throw their weight around and insist that Facebook, Twitter and others provide the desired data sophistication.

Dennis Mortensen

Dennis Mortensen is a pioneer and expert in the analytics industry. Mortensen is an entrepreneur and was COO at IndexTools when it was acquired by Yahoo, becoming Yahoo’s Director of Data Insights. Today he is Founder and CEO of Visual Revenue Inc., sits on the board of the Web Analytics Association, and maintains the highly popular analytics and media blog.

What were the biggest achievements for web analytics in 2010?

We (the industry) finally figured out that this is not about data collection or reporting or displaying up-and-to-right trends. It seems like we’ve come to a point where everybody understands that online data is an opportunity for the organization to make better and more insightful strategic decisions.

What are the most important challenges in 2011 for web analytics and business intelligence?

First and foremost, it is important for those of us in new media not to make a unique distinction between online analysis (web analytics) and that of more traditional business analysis (business intelligence)—as long as we keep the two disciplines separated online comes out on the wrong side of that equation.

My world revolves around online publishers; in particular news media. With that in mind the most important 2011 challenge, as I see it, will be greater data focus and the desire and willingness to be more data driven. Specifically for my target group, editors will be looking to use decision support systems and will be moving towards an environment where machine learned and predicted content recommendations are a daily norm.

Stéphane Hamel

Stéphane Hamel has over 20 years of experience and is the creator of the Online Analytics Maturity Model, WASP and gaAddons. Stéphane is also a frequent speaker at the eMetrics Marketing Optimization Summit and other conferences in North America and Europe.

What do you think were the most important developments in web analytics in 2010?

Market consolidation comes to mind, with IBM purchasing Coremetrics, Unica and a slew of other players in the analytics space, and of course, the previous purchase of Omniture by Adobe. I wrote about this on our blog.

Second, and a topic very close to my interests, is the huge progress of anything related to “online analytics maturity,” as if in 2010, lots of agencies & organizations realized the roadblocks they are facing have nothing to do with the tools, but much more with the organizational shift required to truly entrench a culture of analytics.

What are the most important challenges in 2011 for web analytics?

Stop trying to do everything, do what matters. There is a lot of buzz and coolness factor in the web analytics space. I get crazy when I see anything “social media” receiving a ridiculous amount of attention while businesses are not even able to efficiently help their prospects and clients through their “traditional” online channels (like websites, or even call centers!). We have to wonder which is most important: thousands of friends on Facebook or followers on Twitter… or satisfied and profitable customers. I’m fearing a “social media bubble;” energy and focus is so out of whack that huge deceptions are bound to happen—especially when any attempt to measure the true ROI of social media as failed, turned out to be ridiculously inflated or biased, or as been anecdotal spurs of success at best.

What will be the influence of the “do not track” options now being made available in web browsers?

Did “no not call” bring down the use of phones? Certainly not—it allowed to set the boundaries of what is ethical, legitimate and legal… from blatant abuse. If I’m going to a site I trust and use often, I will gladly accept their offering, including any profiling and targeting. If the “do not track” feature can block LSO Flash objects using features to track users who do not want to be tracked, I’m all for it. If it blocks rogue third party trackers who do not abide by a minimum level of decency, I’m also all for it!

I really would like to know what you guys are thinking. Please add your comments below about what are the most important topics and challenges are emerging in the web analytics and business world.