A few years back, I remember reading a blog post by Avinash Kaushik on the subject of Multiplicity. In theory, I completely agreed with the idea of multiplicity however in practice, I found myself being a laser beam for all things Omniture. I was the perfect case study for the “single source of truth” model. I found myself sitting in strategy meetings where analytics was being discussed and I found myself sitting on the outside, only to be brought in if there was an “Omniture issue” or a question we needed “Omniture to answer”.

It didn’t sink in, that the model I had deployed for myself was against everything I believed in, until I found myself in the mountains, shooting scenes of nature, with two cameras and several lenses. It was then that it hit me, I was doing myself and the company I worked for a disservice by being an “Omniture Shop” rather than a “Web Analytics Shop”. In photography, there is never a single camera or lens or aperture setting that is right for every situation and the same is true for web analytics. If I was going to be a successful analyst, I needed to move outside my Omniture comfort zone and I needed to incorporate more gear into my bag.

Today, I make use of several tools to complete a job. Sometimes, what is required is a high end tool like Omniture Discover, other times something like CrazyEgg will get the job done. And yet others will take a combination of Omniture Discover, Omniture Test & Target, and ProClarity to get the job done right.

It is a mistake, one that I have been guilty of, to think that your high priced analytics tool is the only tool you’ll ever need. If I had made that mistake in my photography, I would have never put down my dSLR and picked up a $12 toy camera from China.

Taken with a $12 Holga camera from China

Jason Thompson

Jason is the Sr. Director of Community Solutions at Keystone Solutions. If you are interested in partnering with Jason at Keystone, in need of analytics & optimization consulting, or looking for employment/internship possibilities, please feel free to email ‘jason [at] keystonesolutions [dot] com’

Twitter - Facebook - More Posts

Even for amateur blog authors, understanding your site traffic is critical to building a site that provides information and services that people are asking for.  Even if you are just blogging about your kids soccer game, understanding where your traffic is coming from and what they are doing on the site can not only prove interesting but may help you target your posts for your most faithful readers.

Google Analytics is by far the most popular free web analytics service that is available but with Yahoo’s acquisition of IndexTools in the summer of 2008, Yahoo has been dipping their toe in the deep end and it looks like they are poised to give Google a run for their money.  If I had to bet one way or another, I’d put my money on Google but I wouldn’t count Yahoo out yet as Google has failed to do anything that would suggest they are ready to step on the throats of their competitors.

I have recently been experimenting with Yahoo Web Analytics (YWA) and have been pleasantly surprised with their offering. For a part-time blogger it is more than powerful enough to give you insight into your site activity.

Recently, Rudi Shumpert, released a WordPress plug-in that automatically deploys YWA tracking to your blog.  You don’t need to know PHP or anything about Yahoo’s tracking code, you simply install the plug-in, provide a few details, and the plug-in does the rest.  I have been testing the plug-in and had it installed and running in under two minutes.

What I love most about the plug-in, is that it is theme independent.  I can change my theme a thousand times and never have to worry about moving my analytics tracking code from theme to theme, the plug-in simply applies to code to all the pages of my blog.  It’s a beautiful thing.

If you have an existing YWA account, installation of the plug-in is easy.  Under the Plugins section in your WordPress admin, click ‘Add New’

Search for ‘YWA’ and click install.

For detailed information about configuration steps and plug-in versions, please visit Rudi’s plug-in site

Jason Thompson

Jason is the Sr. Director of Community Solutions at Keystone Solutions. If you are interested in partnering with Jason at Keystone, in need of analytics & optimization consulting, or looking for employment/internship possibilities, please feel free to email ‘jason [at] keystonesolutions [dot] com’

Twitter - Facebook - More Posts

Hello. My name is Jason. And it’s true, I once worked at Omniture.

It was 3 years ago that I moved on from Omniture to see how things worked on the client side. The company I went to work for had zero web analytics, so I was charged with building the practice from the ground up. I got to work understanding the needs of the business and sending out RFPs. Although I had come from Omniture, I was keeping my companies best intest in mind, and I was set on choosing the vendor that would provide us the greatest long term value.

However one vendor decided that I was simply wasting everyones time, including my own, by going through the RFP process. I was so entertained by this voicemail, that I had it converted to MP3 and had filed it away.

Vendor_Call

One thing I have learned working on the client side is that there never is a single tool that answers all your questions.  If you take a peek under the covers, you’ll notice that I’m running Omniture, Google Analytics, and Yahoo Analytics on my blog, I’m an equal opporunity analyst.

Jason Thompson

Jason is the Sr. Director of Community Solutions at Keystone Solutions. If you are interested in partnering with Jason at Keystone, in need of analytics & optimization consulting, or looking for employment/internship possibilities, please feel free to email ‘jason [at] keystonesolutions [dot] com’

Twitter - Facebook - More Posts

We tend to gravitate to the sensational, hence why nightly news shows are filled with stories of rape and murder and conspiracy, we know deep down it’s wrong, yet we can’t look away. When we get sucked into this world of mystery and intrigue, we tend to throw out logic and believe anything that is flashed before our eyes.

Welcome 2012. The year the world will end. Well, according to popular Mayan belief. But was it really their belief that the world would end on December 21, 2012?

Popular theories of the Mayan’s prediction of the end of the world are based on a very limited set of artifacts. Many believe the Mayans predicted the end of the world based on evidence found in a single book and a set of stone discs. What if the set of discs was incomplete? What if the book was part of a larger volume?

Is this a fair representation of what the Mayan culture believed? Imagine if 10,000 years from now, archeologists discover a single book, containing the story of Hansel and Gretel, while unearthing a public library in Upstate New York. Would it be fair to say that, based on this finding, people that lived circa 1800-2500 lived in houses made purely of a sugar based substance, sometimes baked, other times not, and whose children were susceptible to being eaten by hunched over old ladies wearing over-sized black hats?

Yet, we do this all the time when we analyze the behavior of our web site visitors. We take a nugget of information, a singular fact, and from that we generalize a grand explanation for how an entire system works. I have been guilty of this way of thinking many times.

If we can take one lesson from the apocalyptic hype surrounding the year 2012, it is that we as analysts should be careful about getting sucked into the emotion of sensational news. Rather than becoming overly excited about the first artifact we uncover, we should center ourselves and logically look at all the information we have available to us. We should continue to dig for more truths. And, we should be slow to profess universal truths based on a limited set of data.

Jason Thompson

Jason is the Sr. Director of Community Solutions at Keystone Solutions. If you are interested in partnering with Jason at Keystone, in need of analytics & optimization consulting, or looking for employment/internship possibilities, please feel free to email ‘jason [at] keystonesolutions [dot] com’

Twitter - Facebook - More Posts

A. The percentages in column A represent the conversion rate for each step relative to the entry point into the funnel. In the example, the 1.19% conversion rate for ‘Compose Message’ is calculated as ‘Visits to Message Compose / Visits to Central’ — in the context of the path: ‘Central’ to ‘Advanced Search’ to ‘Compose Message’.

B. The first conversion metric is the lowest daily conversion % between the two steps, in the current reporting period.

C. The second conversion metric is the highest daily conversion % between the two steps, in the current reporting period.

D. Visit conversion rate between the two steps.

E. This is the fall-out percentage between steps.

F. This shows a ranked view of the top attrition pages between steps.

Jason Thompson

Jason is the Sr. Director of Community Solutions at Keystone Solutions. If you are interested in partnering with Jason at Keystone, in need of analytics & optimization consulting, or looking for employment/internship possibilities, please feel free to email ‘jason [at] keystonesolutions [dot] com’

Twitter - Facebook - More Posts

Did I catch your attention? Email subject lines are a critical part of any good marketing plan and should be tested, tested, tested. However, I fear the conversion funnel of Send–>Open–>Click–>Conversion is typically overlooked in our fast paced, get it done and out the door now, world of online business.

I recently received an email from Threadsy, touting the new features of their “all-in-one social messaging aggregator”. Typically emails from vendors go straight into the trash but this one was a little bit different. See the email below:

Now, I find it hard to believe that the subject was just a random array of words, rather I believe the clever marketing team at Threadsy is using the age old adage that “sex sells” and it sold me. “More bang for your box?”  Yes, I opened.

I’m really hoping that Threadsy ran an A|B test on this email campaign, mostly because I’d be really interested in seeing the results.  Anyone over at Threadsy want to email me, I’d love to talk about the results, even at a high-level.

Q67PG9D3MU2Q

Jason Thompson

Jason is the Sr. Director of Community Solutions at Keystone Solutions. If you are interested in partnering with Jason at Keystone, in need of analytics & optimization consulting, or looking for employment/internship possibilities, please feel free to email ‘jason [at] keystonesolutions [dot] com’

Twitter - Facebook - More Posts

Version 0.4 of the Omniture Twitter Analytics Plug-In is now available. The new version offers a simplified way of tracking multiple keywords.

Tracking Multiple Keywords in v0.3

$q = “+omniture+OR+omtr+OR+adobe”;

Tracking Multiple Keywords in v0.4

$q = “omniture,omtr,adobe”;

You no longer are required to know the exact syntax of the Twitter search, simply add a list of comma separated keywords and the plug-in will format the search string for you.  Although this seems like a simple enhancement, I have received numerous requests about how difficult it was to track multiple keywords, so I’m hoping this makes things much easier.

If you are an existing user of the Omniture Twitter Analytics plug-in user (v.03 or older), email me at ‘jason [at] omnituretwitteranalytics.com’ and I’ll send you a download link and instructions for how to deploy the updated code base.

If you are a new user of the Omniture Twitter Analytics plug-in, head over to our product site and request the application.

Jason Thompson

Jason is the Sr. Director of Community Solutions at Keystone Solutions. If you are interested in partnering with Jason at Keystone, in need of analytics & optimization consulting, or looking for employment/internship possibilities, please feel free to email ‘jason [at] keystonesolutions [dot] com’

Twitter - Facebook - More Posts

Step 1: Update your s_code.js file with the Omniture TnT Integration Plugin – Version 1.0

/*
* TNT Integration Plugin v1.0
*/
s.trackTNT = new Function("v", "p", "b", ""
+ "var s=this,n='s_tnt',p=p?p:n,v=v?v:n,r='',pm=false,b=b?b:true;if(s."
+ "getQueryParam){pm=s.getQueryParam(p);}if(pm){r+=(pm+',');}if(s.wd[v"
+ "]!=undefined){r+=s.wd[v];}if(b){s.wd[v]='';}return r;");

Step 2: Determine if you will include the integration call in your standard SiteCatalyst page tag or if you will require to use a Custom Link call.

Is your SiteCatalyst code block located after the last mBox call on your page?

YES!  Simply add the following before you call s.t():

var tntInput = s.trackTNT();

s.eVar47 = s.tnt = tntInput;

Is your SiteCatalyst code block located before the last mBox call on your page?

YES! This will take a bit more work but, if you ask me, worth it.  I have our SiteCatalyst code directly following the <body> tag.  This has greatly improved our data accuracy.  Moving the code to accommodate an integration with Test & Target was simply out of the question.

To send the Test & Target information to SiteCatalyst using a Custom Link call add the following after all mBoxes on the page:

<script>
var tntInput = s.trackTNT();

trackTNT(s_account, tntInput);

function trackTNT(s_account, tntInput){
	var s=s_gi( s_account );

	s.linkTrackVars="tnt,eVarN"
	s.linkTrackEvents="None"

	//variable for TNT classifications
	s.eVarN = s.tnt = tntInput;

	s.tl( true , 'o' , 'For tracking TNT' );
	}
</script>

Step 3: Classify the Test & Target Campaign and Recipe ID. Omniture has a behind the scene integration to automatically classify your TnT integration. However, I have been told it is limited to one (1) report suite. I found it easy enough to simply do this myself.

Step 3a: Create a Custom Conversion variable to capture your Test & Target integration IDs

Step 3b: Classify your conversion variable to include a text based classification called ‘Treatment’

Step 3c: This part is a little hacky but hey, it works.  Log into Test & Target and view your campaign.  The TnT plugin will return a value like ’4503:0:0,’.  To get the values hover your mouse over the preview icon for each treatment

Now, look down in your status bar and you will be able to see the values you need to write down.

Now you can compile a list of Ids and treatment names

4503:0:0,       Control

4503:1:0,        Improved Headline

Step 3d: Download a SAINT template for your TnT classification.  Map each treatment id to your friendly treatment name.  Upload.

Step 4: Create and activate a plug-in offer in Test & Target.  This plug-in will pass the Test & Target identifiers to the page.  For detailed instructions on creating this offer, refer to ‘Configuring the SiteCatalyst Integration’ in Test & Target online help.

Step 5: Analyze your results.  Test & Target is great for giving you a dashboard view of your treatments and how each one is converting (with statistical significance).  However, you still are not able to fully understand how your tests are changing user behavior inside your site.  This is where SiteCatalyst and more powerfully, Discover, comes into play.  Use SiteCatalyst and Discover to slice your visitor population to better understand how your Test & Target campaigns are influencing user behavior beyond the conversion check point.

And by uploading a friendly treatment name via SAINT, this job is made much easier.

Jason Thompson

Jason is the Sr. Director of Community Solutions at Keystone Solutions. If you are interested in partnering with Jason at Keystone, in need of analytics & optimization consulting, or looking for employment/internship possibilities, please feel free to email ‘jason [at] keystonesolutions [dot] com’

Twitter - Facebook - More Posts