In Part 1 of this series I explained link tagging, the technology that Google Analytics uses to track on-line marketing campaigns. In Part 2 I discussed how to tag your links and posted a tool that I use to quickly tag large numbers of marketing URLs. Today, in Part 3, I’ll start to pull this whole thing together by walking through a very basic analysis.
How I Start
I like to start my marketing analysis using the Marketing Campaign Results reports. Using these reports I can immediately identify any campaigns that are under or over performing. They’re a great launching pad for further analysis. You can find them in the Marketing Optimization > Marketing Campaign Results section.
The reports segment the data based on where the visitor came from using the values from the campaign tracking variables. So, for each of the major campaign variables we discussed in Part 1 (utm_campaign, utm_source and utm_medium) we have a corresponding Google Analytics report. Here’s the mapping of campaign variable to GA report:
|Campaign Variable||Report Name|
This means that the values you used in the campaign variables will be pulled directly into the reports. Exciting stuff, huh? :)
Campaign Conversion Report
Let’s start with the Campaign Conversion report.
This report segments the traffic based on campaign name. It contains information from tagged URLs (using the utm_campaign variable) and un-tagged URLs. How does it get data for the un-tagged URLs? If you’re using the auto-tagging feature in AdWords then Google Analytics will automatically pull in the Campaign names you create in AdWords. All other un-tagged URLs get put into the following buckets:
- (direct): visitors that entered your website address directly into the browser
- (organic): visitors from an un-paid search engine listing
- (referral) : visitors that clicked on an un-tagged link
- (not set) visitors from links that were tagged but were missing some information. For example, if you are looking at the Campaign Conversion report, and see that there were 10 visits from ‘(not set)’ this means that the utm_campaign variable was missing from the tagged link.
Ok, so what does this report tell us? It helps us quickly understand how well our campaign is performing using some basic metrics:
- Visits: How much interest did the campaign generate?
- Goal Conversion Rate (G1/Visits): Did the visitors from this campaign do what we wanted them to do?
- Transaction Average (T/Visits): How many transactions were generated by this campaign?
- Revenue per Visit ($/Visits): How much money did we make from each visit in the campaign?
It’s important to realize that each metric gives you a bit more insight into what is going on. For example, let’s say a campaign has a very low conversion rate. Why? Look at the number of visits. Is the campaign generating a lot of traffic? If there are a high number of visits but a low conversion rate there may be a disconnect between the marketing message you’re sending and the content the visitor sees when they land on the site. Dig deeper, Try checking the bounce rate for the landing page.
Again, this is a good starting point for a deeper analysis. And analysis means segmenting the data to gain more insight.
Segmenting Campaign Data
Notice that the first column in the report is named ‘Campaign/Source’ and not just ‘Campaign’? The reason is that this report let’s us drill down into our campaign and view the sources associated with the campaign. If we click on the plus sign for the ‘Ongoing’ campaign we can drill into the data and see the associated sources.
This tells us is there were three sources of traffic in the ‘Ongoing’ campaign: Squidoo, Wikipedia and MySpace. This is real data from a company using social networking and viral sites to drive traffic. The value in the brackets (Social_Networking and viral) is the medium (which we’ll get to later).
Remember from Part 1 of this series, the source is the ‘who’ part of the campaign. ‘Who’ did we partner with to distribute our message? By drilling down into the data we can find out. Drilling from the campaign level to the source level revealed a lot about our campaign. It looks like this business should dump MySpace and focus more on Squidoo! Not only do they get more traffic, they get far better conversion.
But let’s take a different look. It could be that these sources are used in multiple campaigns. Maybe MySpace did really well in a different campaign. Let’s use the Source Conversion Report to get a different view.
Source Conversion Report
The source report shows us how all of our various sources are doing. It does not matter which campaign the source belongs to, they are all listed in this report. This report is very helpful because it shows, historically, how well a source performs. It may be that a source just under performed for a specific campaign. We can see above that Squidoo, which performed very well in the Ongoing campaign, does not crack the top 10 sources. MySpace is no where to be found. This probably means that MySpace performs poorly across the board, not just in the Ongoing campaign.
Next to each source we can see the medium associated with that source. Again, I like to think of the medium as the mechanism that we use to push our marketing message out. Was it email, CPC, banner, print, etc. Google Analytics has pulled the medium value from the utm_medium variable and placed it in the report.
Medium Conversion Report
Looking at the medium we can evaluate how well the mechanism is working for us. Let’s see how well the Social Networking ‘mechanism’ is working.
Interesting. We can see that the ‘Social Networking’ medium doesn’t get a lot of traffic, but it gets attentive traffic (high number of pageviews per visit) and what appears to be an average conversion rate (for this site) for Goal 1.
The medium report is good at identifying dependencies. Are you too dependent on a particular way of getting traffic? If all your conversions come from organic, and the search engines drastically change their ranking algorithms, then you could loose a lot of traffic and a lot of money!
In addition to the Conversion reports, there are also three ROI reports. These reports are very similar to the Conversion reports. They segment the data in the same way (based on campaign, medium and source). The difference is the metrics reported. Rather than conversion rates, these reports show cost, revenue and ROI. If you have an e-commerce site and are collecting revenue, or have monetized the values of your goals, then the revenue generated by each campaign will be displayed.
A warning about this report. GA will only pull in cost data from your AdWords campaigns. Do not be alarmed if you see no other cost data in this report. GA is a closed system, you can not import cost data from other sources. This means that the ROI calculations will be incottrct for non-AdWords campaigns.
The referral conversion report is another fantastic report :) This report lists all of the un-tagged, non-organic and non-direct links that drove traffic to the site.
Drilling down into this report will show you where on each referral domain, the visitor originated. I find the referral report enlightening. The web is a wacky place. And people reference content in so many different ways. This report will help you hunt down all the sources of your traffic.
Some Final Thoughts
You may notice that some of the reports above have multiple lines for the same items. For example, the Medium Conversion report has two line items for social networking:
The reason the item is listed twice is that the person tagging the links specified two different values for the utm_medium variable. That’s why it’s important to use a standard naming convention when tagging your links.
Well, that wraps up this Part 3 of our Campaign Tracking series. What did you think? My ultimate goal is to make you all marketing measurement wizards. Am I doing a good job?
Daniel Waisberg says
Very insightful! But I have a question, one problem I have been working on for a while regarding analyzing Adwords campaigns.
The “Campaign conversion” report you showed on this post do not drill down for adgroups and keywords (at least for me). And the “Adwords analysis” report on GA has only the information you already have on Adwords, such as: impressions, clicks, Cost, revenue, CTR, CPC…
However, what interests me the most is P/visit, and Goals’ conversions for keywords. But my campaigns are divided between GEO locations, and I usually have the same keywords for different regions (which mean I cannot simply use the “Overall Keyword conversion” report, since it is not divided by campaigns). I could cross-segment each keyword (which would be a nightmare), but Google saved me, since they do not allow me to do that because my data is ‘too big to cross segment’ according to them.
The only solution I found was to create a profile for US, for example, and use the “Overall Keyword conversion” for CPC, which I can manually link to my US campaigns (keywords appear only once per country).
Well, I feel there might be an easier solution (I would prefer to see all the information in one profile). What do you think?
Thank you and keep on writing; your blog is a great source of knowledge for GA users and Web Analysts in general…
First, thanks for the words of encouragement. Sometimes it’s tough to if this information is helpful. Comments like yours are validation.
You pose a very interesting question. I would try using an advanced filter to concatenate the campaign name to the keyword. Something like this:
“campaign_name – keyword”
Then, when you use the Overall Keyword Conversion report the data will show both the campaign name and the keyword all in one line.
A word of warning, I’m not 100% sure if this will break the conversion tracking for each specific keyword. My advice is to try the filter on a test profile.
Thanks and good luck!
This article was super clear and easy to understand. I could have guessed that your job is to educate people with analytics ;-)
Luke Humble says
Again, thanks for this article as it was very clear and easy to understand.
There is one area that l am interested in obtaining extra information on as there seems to be a lot of talk regarding this on the GA Group – (not set) data.
You mentioned this in your article but l am interested to know if there are any other possibilities for why l user would experience the (not set) data appearing in their results?
I’m about to track a campaign using the link tagging but is it ok for the value of the utm_term to be keyword groups and not individual keywords?
Our media buyers are suggesting this would be much more manageable for them, rather then give then individual tracking links for each keyword…?…
Justin Cutroni says
Yes, you can use utm_term to track keyword groups rather than specific keywords. In fact, you can use utm_term to track anything. They key is that you use utm_term to track actionable data.
Good luck with your campaign,
Chris G. says
I am trying to setup tracking for a banner ad we have on a third party site. Setting up the link is easy. Finding the results is my stumbling block. You say to go to the Marketing Optimization > Marketing Campaign Results section. That’s my problem, I don’t see that in GA, which is frustrating because you walk us through the process step by step. Am I blind? Have the names been changed since you wrote this article? Any help is much appreciated.
Justin Cutroni says
Sorry for any confusion that this post has caused, it is a bit out of date.
When Google changed the GA UI it moved the reports around. You can ow found the information about your banner ad in Traffic Sources section. The All Traffic Sources report will show data originating from your banner according the source and medium you specify in the link tags.
You can also use the Campaigns report, found in the Traffic Sources section, to identify traffic from your banner based on the utm_campaign value.
Hope that helps and I’ll update the post soon.
Very informative post. However, I can’t seem to follow your demonstration on how to drill down. In the new interface, Google let us look at campaigns via source, medium, content etc… but doesn’t seem to allow us to cross reference them. For example, how does one look at the traffic due to different medium due to a single campaign name? I could be missing something, but without this ability to drill down, the utility of this is seriously limited.
Justin Cutroni says
When Google changed the user interface in October, 2007 it completely broke this post. Unfortunately I have not had a chance to update this, and a few other posts, that are out of date. Sorry!
You ca still drill down into the data using segmentation, it’s just a bit different now.
Start by navigating to the Campaigns report and then click on a campaign. The resulting screen will show all sorts of info for that screen. Notice under the graph that there is a segment drop down box? You can use that box to segment the data by other attributes like medium or source.
Thanks for reading and I hope that answers your question.
Can you tell me where to locate the campaign conversion report in analytics? And also where to place the tag links? Anywhere in the body tag?
Justin Cutroni says
You can find the campaign conversions in the Traffic Sources > Campaigns report. Then click on the Goal Conversions tab. GA will segment all conversions based on campaign.
Thanks for the question and best of luck measuring your campaigns!
What if a visitor (1) Arrives by clicking a Google organic search result, then, (2) Comes back the next day by clicking on an AdWords ad, then (3) Then comes back the next day by a new Google organic search result, and this time they arrive at a Goal page so a conversion is recorded.
But what will happen within Google Analytics? Will the conversion show up as a result of the last (organic) visit? Will there be no AdWords conversion showing in Google Analytics because Google Analytics keeps updating the user’s cookies and the conversion happened on the third (organic) visit?
Thanks very much if you know the answer to this.
Justin Cutroni says
In Google Analytics you’ll see the following:
1 Visit from Google Organic
1 Visit from Google AdWords
1 Visit from Google Organic (2nd search) WITH conversion
There will be no conversion attributed to Google AdWords. Google Analytics uses a last click attribution model meaning the last click gets credit for the conversion. The only exception is unless the last click is a DIRECT visit to the site. Direct traffic will not overwrite the existing campaign information.
Hope that clarifies things.
I’m extremely interested in learning how to effectively use GA to track twitter leads to my company’s website. But I’m having trouble finding the Marketing Optimization section that you refer to. Is your information still up to date? Are these reports generated from within GA? I’ve found the information you posted here EXTREMELY informational, interesting and helpful otherwise. Thanks so much. ~Dinah
Justin Cutroni says
Try this post on how to track Twitter.
Great, post, I love your content and topics, they are very well written. I have purchased your book and plan to attend one of your “Seminars For Success” conferences..
I’m wondering how I can track which referrals my ad is appearing on? I have the same ad appearing on over 100 sites through a media buyer so I’ve given the buyer 1 tracking link that would look like this:
I’m wondering if its possible to track which sites this ad has appeared on. I’m thinking this would help me later with the media buyer decide which sites perform well and which don’t. Currently I see in GA that I can only filter to the campaign, but not find the referral.
Thanks in advance!
Justin Cutroni says
@Mike – You could create 100 different links, and change the value of utm_source to be the domain of the site where the ad is appearing. That would work, but it’s very manual. Or you can talk to the person trafficing the ads and ask them if there is some type of syntax that will dynamically add the name of the site to the destination URL. For example, some ad platforms will let you do something like this:
And the site name will automatically be pulled into the URL. That’s the best solution. But if that does not work you can create 100 different links.