Why Omniture
This is a good overview video of what it is and what it’s useful:
http://www.omniture.com/product_tours/sitecatalyst
Presents information in a way that is useful and meaningful to your specific company. Can customize, share and notify people on a standard schedule.
You can see the value each of your page’s has. As people move through the site, you can gauge the value of each page they visit based on your overall most wanted response.
*******************
I’m going to be using this page to accumulate notes for Omniture.
(Sorry my blog is so freakin’ boring these days)
Basically, if you want to tie your data to traffic metrics, use a prop. If you want to tie your data to your site?s custom success events, use an eVar.
There are many options for page naming. If your site is organized well, they PageName plugin might be the way to go. The code can be found in I:\PPC\Omniture\PageName plugin variables – SageRock 2006.txt
It goes in the .js file. If the site is dynamic server-side logic is probably the best way to go. You can read about it in the book on page 17 of the Advanced Implementation manual.
These training modules are recommended:
SiteCatalyst New User Orientation
Commerce and Conversions
Custom Variables and Events
Introduction to Implementation
Implementation Strategies
******************************
Configuration Variables:
s.channel – this would be used for Site Sections. Each page of a site should probably have this assigned.
s.prop – Custom Insight Variables. These are counters. They can show content type, templates or subsections.
To Add or Edit s.prop names, go to Admin: Report Suite Manager: Edit Settings: Traffic: Custom Insight Variables
s.events
Predefined Events:
prodView – Success event occurs any time a visitor views a product.
scView – Success event occurs any time a shopping cart is viewed.
scOpen = Success event occurs any time a visitor opens a shopping cart for the first time.
scAdd – Success event occurs any time a product is added to a shopping cart.
scRemove – Success event occurs any time an item is taken out of a shopping cart.
scCheckout – Success event occurs on the first page of a checkout.
purchase – Success event occurs on the final page of a checkout (includes Revenue, Orders, and Units).
So:
s.events=”prodView”
s.events=”scView”
s.events=”scOpen”
s.events=”scAdd”
s.events=”scRemove”
s.events=”scCheckout”
Then there are custom events:
s.events=”event1″
s.events=”event2″ etc…
How do I assign friendly names to my Custom Events? (s.events)
Using SiteCatalyst, you can change the names of your custom event variables. Changing the name of the custom events will not have an impact on the web site code or the data supplied to or reported by SiteCatalyst. Instead, the report will display the friendly names you entered in place of the default custom event variable names, i.e. Custom Event 1. The steps below enable you to assign a friendly name to your custom events.
1. Log in to SiteCatalyst.
2. Click the Admin tab.
3. Click the Report Suites tab.
4. Select a report suite.
5. Click Edit Settings.
6. Click Conversion, and select Success Events.
7. Click the check box next to the desired success event to select it.
8. Type the friendly name.
9. Click Save.
s.eVar – Custom Commerce Insight Variables. Used to segment customer success.
To Add or Edit eVars go to Admin: Report Suite Manager: Edit Settings: Conversion: Conversion Variables. *I think* each evar is a success event… i.e., newsletter sign up, form filled out. And then when goes in the variable, on the page, are the details of that. You get 10 of these. But you can buy up to 50. Each one can be 100 characters.
**********************
Custom Traffic Variables can be used to have multiple page names for a single page based on how a person has pathed to a page.
***********************
Commerce and Conversions
Types of web analysts:
Operators – Do very little analytics. Get about 3% conversion rate.
Tacticians – Looks deeper in numbers. Get about 5% conversion rate.
Strategists – Go very deep with recommendations. 7% conversion rate.
Strategic Process:
First, must define Business Goals. Typically can be broken down into smaller goals.
Second, think about the key performance indicators that determine if you are reaching those Business Goals. This could be revenue on the site or revenue per visitor.
Third, Tactical metrics. Use site catalysts that will give you the numbers to report on your KPI.
Conversion is any process that your would like a customer to complete: Look at specific content, site registrations, individual product sales, marketing campaigns, loyalty, etc. A “customer” is a person that is doing the “thing” we want them to do.
**********************
White Paper – Multivariate Testing (MVT)
https://omniture.custhelp.com/cgi-bin/omniture.cfg/php/enduser/std_adp.php?p_faqid=680&p_created=1127507331&p_sid=zNh_XTCi&p_accessibility=0&p_lva=&p_sp=cF9zcmNoPTEmcF9zb3J0X2J5PWRmbHQmcF9ncmlkc29ydD0mcF9yb3dfY250PTI5JnBfcHJvZHM9JnBfY2F0cz0mcF9wdj0mcF9jdj0mcF9zZWFyY2hfdHlwZT1hbnN3ZXJzLnNlYXJjaF9ubCZwX3BhZ2U9MSZwX3NlYXJjaF90ZXh0PWxhbmRpbmcgcGFnZXM*&p_li=cF91c2VyaWQ9c3RkdXNlciZwX3Bhc3N3ZD0mcF9saV9wYXNzd2Q9bXljZXh0&p_topview=1
Taguchi Method
http://en.wikipedia.org/wiki/Taguchi_methods
As you design your experiment, you need to keep your overall goals in mind.
? State your objective.
? What do you want to improve (the key metric)? This metric is how you will gauge the influence of the various page elements tested in the experiment.
? What do you hypothesize will help you improve? Identify the regions or attributes on your page (your Page Elements) that you feel will influence your key metric. For each of these elements, you need to identify both the current and the new version.
Note that there are two sub-phases in running the experiment (Phase 2):
o Getting the baseline with placeholder recipes
o Getting data for the actual recipes
Phase 1:
Focus your experiment on one of the phenomena that you?ve noticed by summarizing what you want to find out. This is your experiment ?objective.?
Select Your Key Metric – Dollar per Order or Conversion Rate
Not a Flat Metric: Cart Adds or Cart Opens
Describe Page Elements – Identify the regions or attributes (?Page Elements?) on your page that you feel will influence your Key Metric. These are essentially your experiment ?variables?. For each Page Element, identify the change (?New Version?) that you hypothesize will improve your Key Metric
Start with no more than 7 page elements for your initial test.
SiteCatalyst will provide a PDF which describes your Experiment Design.
Suggested Test Items
Design, Layout, Navigation
* Branding vs. selling imagery
* Copy styles
* Merchandising ? what products should I show on my landing page?
* Image sizes ? do I need to show big pictures of my key featured product? Or is it better to show mall pictures of various products?
Promotions
* Promotion targets
* Promotion levels
* Pricing levels
Phase 2: Implementing the Experiment
Designate an eVar for use in tracking your experiment. Your IT team will assign the recipe name to this eVar in the web beacon tracking code, so that you can attribute metrics to each recipe in your experiment.
Tagging pages:
Each “recipe” is a letter. If you have 8 recipes they will be A-H
Make sure you do not add the word ?Recipe? to the eVar value.
Set up a Test Harness and Deploy Recipes
In order to test your recipes concurrently, you will need to segment your traffic into equal testing groups (?Visitor Test Groups?). This segmenting mechanism (?Test Harness?) can be challenging to calibrate accurately. The number of visitor test groups should match the number of recipes that SiteCatalyst has described.
Sample Size:
Typically, you need between 50 to 100 instances (minimum) of your key event in order to achieve confidence in the results.
You get to the MVT section by clicking on “Advanced Analysis” and then “Multivariate Testing Framework” at the bottom. Watch the “Design your experiment” section of the MVT video.
Paid Search Reporting
You need to update your global js file to activate this:
In our implementation, we actually have a rule in our global JS file that looks for a certain tracking parameter (in your case it would look for the word “source”), and if that parameter is passed through the URL from Google or Overture, the JS file will write the code into the s_campaign variable, which will pass into SiteCatalyst.
*****************
Material I have read or watched:
Implementation Strategies Video
Multivariant whitepaper and video
Commerce and Conversions Video