This chapter covers the following topics:
Today’s e-commerce environment necessitates that companies measure and understand online visitor interactions and customer purchasing behavior in terms of the impact of these activities on the revenue and profitability of their Web initiatives. Oracle® Web Analytics addresses this need by empowering managers with insights to deliver superior Internet campaigns and drive profitable customer interactions. Business managers are required to run their online businesses like their offline business. They need tools to answer tough questions on adoption and ROI for their online initiatives such as:
Which sources bring the most customers, orders and revenue? What is my conversion funnel?
Which navigational paths lead to the most conversions?
Is my online channel profitable?
Does my online channel have usability issues slowing down the adoption?
How are my summary traffic statistics (visitors, pages, visits, etc.) changing over time?
How can I leverage my customer analytics from the Web in my multi-channel process?
An integrated part of the Oracle E-Business Suite, Oracle Web Analytics answers these questions and more. Oracle Web Analytics is a flexible and robust tool for tracking and reporting on site activity. A comprehensive reporting system, Web Analytics can be used in both business-to business (B2B) and business-to-consumer (B2C) settings. With seamless integration to the e-commerce application, Oracle iStore, Oracle Web Analytics provides a complete Web tracking solution, reporting on the entire online customer flow, from clicks, to registration, to leads, to orders. Oracle Web Analytics, along with Oracle Marketing, provides marketing managers with a complete Web campaign optimization solution. By monitoring which Web campaigns translate into site visits, registrations, and revenue, Oracle Web Analytics is able to improve the ROI of marketing campaigns.
Oracle Web Analytics targets the following business roles:
Web Channel Business Owner: The online sales business owner is responsible for the sales and customer activity on the site devoted to capturing purchases. This person may typically assume the title of Director of E-Business, with responsibility for an organization’s e-commerce initiatives and its operational efficiency therein. In a B2B context, the individual would belong to a sales organization, while in a B2C scenario, the individual would typically report within a marketing organization.
E-Marketing Manager: The e-marketing manager plays a key role in B2C Web environments: she/he is typically responsible for launching and monitoring web campaigns, as well as for the overall customer online experience. Her/His role would typically be the primary one in this type of organization. This function, however, depending on the size and business objectives of the organization, may not exist in a B2B scenario.
Usability Specialist: The Usability Specialist would focus on specific transaction flows in a Web site to ensure optimal purchasing flow. Examples of typical flows are the checkout and registration.
System Administrator: The System Administrator (also known as Operations Manager) would utilize Web traffic data to assist in scheduling optimal times to disable the site in order to perform maintenance or upgrades.
Oracle Web Analytics captures in a scalable way online users’ behavior and interactions. Following are key features of the Web tracking infrastructure:
Web Tracking Functionality
Web Analytics Administration Application
Privacy Management
Security
Support for Additional Deployment Options
Oracle Web Analytics ships with out-of-the-box tracking of Oracle iStore sites, and can be configured to track legacy third-party sites.
These features are described in the sections that follow.
Oracle Web Analytics provides an extensive tracking infrastructure, capturing customer browsing patterns and Web interactions, such as site registration, site searches, and online inquiries. Following are some key metrics being tracked by Web Analytics:
Visits: Each visit is tracked, and a configurable time-out mechanism defines the visit time-out period.
Visitors: A unique identifying cookie tracks each visitor. In addition, anonymous visitors are tracked across multiple visits.
Page Views: Each individual page displayed to the customer is tracked, easily capturing the users' browsing patterns.
Web Referrals: Advertisements or portal links on partner sites are counted.
Site Registrations: Registration submissions are tracked, along with the visitors’ browsing patterns prior to registration.
Cart and Online Order Creation: Cart and order creation are tracked, in order to analyze which visits and visitors are converted into actual purchases.
Online Inquiries: Payment, order, and invoice inquiries are tracked.
Important: Oracle Web Analytics does not track visitors who have disabled browser cookies. For more information on how Oracle Web Analytics treats browsers with JavaScript disabled, refer to the section "Tracking Visits for Browsers With JavaScript Disabled", below.
Oracle Web Analytics tracks authenticated visitors persistently across visits, regardless of the Visit Time Out value in the administration setup, unless of course the visitor has selected to opt out. If at any time during a visit, a visitor authenticates himself, the entire visit is considered authenticated. The determination of whether a visitor is a new or return visitor is based using the values obtained from the authenticated visitor state. The scenario below helps illustrate the behavior.
| Visit | Visitor ID | Party ID | Tracked |
|---|---|---|---|
| 1 | A | Anonymous | Start of Visit |
| 1 | A | X | At the end of visit 1, the visit is tracked as an authenticated visit made by new Party X |
| 2 | B | Anonymous | At the end of visit 2, the visit is tracked as an anonymous visit made by new visitor B |
| 3 | B | Anonymous | Start of Visit |
| 3 | B | X | At the end of visit 3, the visit is tracked as an authenticated visit made by returning Party X |
| 4 | B | Anonymous | At the end of visit 4, the visit is tracked as an anonymous visit made by returning Visitor B |
Analysis:
Number of Visits: 4
Number of Visitors: 2
Although we know from the sequence of events that Visitor B is really the same as Visitor A, and that there is probably only one visitor tied to Party X involved in the above sequence, in order to protect the privacy of the visitor Oracle Web Analytics does not relate the information across visits.
Oracle Web Analytics tracks anonymous visitors for a period defined in days, as specified by the user in the administrator setup parameter, Persistent Cookie Expiration, unless an anonymous user opt-out is present. The expiration date for the tracking of anonymous visitors is re-set on every visit. For example if the visitor expiration is set to be 90 days, the visitor will be tracked as a new visitor if his next visit is 91 days after the original visit. If the visitor should return before the 90 day period, the visitor will be considered a return visitor, and the expiration date will be re-set for another 90 days.
Administrators can disable persistent cookies in the Administration UI. This, in effect, disables the tracking of anonymous visitors across visits. Thus, anonymous visitors are counted as distinct visitors with every new visit. The following scenarios help illustrate the behavior.
| Visit | Visitor ID | Party ID | Behavior |
|---|---|---|---|
| 1 | A | Anonymous | At the end of visit 1, the visit is tracked as an anonymous visit made by new Visitor A |
| 2 | A | Anonymous | At the end of visit 2, the visit is tracked as made by a new, anonymous visitor even though it was made by Visitor A |
| 3 | A | X | At the end of visit 3, the visit is tracked as an authenticated visit made by new Party X |
| User shuts down browser, and clears session cookies. Visitor then re-starts browser, and continues with visit. | |||
| 4 | B | Anonymous | At the end of visit 4, the visit is tracked as an anonymous visit made by new Visitor B |
| 5 | B | X | At the end of visit 5, the visit is tracked as an authenticated visit made by returning Party X |
Analysis:
Number of Visits: 5
Number of Visitors: 3
Although we know from the sequence of events that Visitor B is really the same as Visitor A, and that there is probably only one visitor tied to Party X involved in the above sequence, in order to protect the privacy of the visitor Oracle Web Analytics does not relate the information across visits.
Note that the above scenario also applies to the following section, "Tracking Visitors Who Disable Persistent Cookies", since the effect of not using persistent cookies is the same -- regardless whether the cookies are disabled in the Oracle Web Analytics setup or on the client machine.
Visitors may browse a site with persistent cookies disabled in their browsers. Oracle Web Analytics supports this browser configuration and continues to track such visitors and visits by setting only session cookies. When session cookies are used, an anonymous visitor is considered a new visitor in every new visit where the session cookies have been removed.
When a user has disabled JavaScript in his browser, following is the tracking behavior:
Tracking External Web Sites (Non-Oracle iStore)
External Web sites are tracked based on the JavaScript code published in their Web pages. Thus, when JavaScript is disabled, no external sites will be tracked.
Tracking Oracle iStore/Oracle E-Business Suite Applications
When Oracle Application Server Web Cache (Web Cache) is Not Enabled: Page tracking of Oracle iStore pages are handled by Oracle Web Analytics APIs, which are JavaScript-independent. Thus, they will be tracked normally. However the Page Duration is computed based on JavaScript code, so the duration of pages tracked on JavaScript-disabled browsers will be computed based on "Extended Calculation" methodology.
When Oracle Web Cache is Enabled: Web-cached pages (for example, catalog pages) will not be tracked. This includes the partially Web-cached iStore pages as well. Non Web-cached pages (for example, order tracker pages) will be tracked. Thus, if JavaScript is disabled, Oracle Web Analytics tracks visits partially depending on page visited by the visitor. As in the above case, Page Duration will always be computed based on "Extended Calculation" methodology.
For tracking purposes, Oracle Web Analytics supports Microsoft Internet Explorer Version 5.5x, 5.0x and 6.0; AOL Netscape Version 4.7x and 7.2; and Mozilla Firefox Version 1.7 (where x is 3 or higher). Browsers with a lower version from the type described above or of a different type (e.g., Opera) are still tracked. However, if the JavaScript code fails to execute on these un-supported browsers, the visits will not be captured. Browsers of a higher version from the types described above are treated normally.
Oracle Web Analytics features a seamless integration with the Oracle iStore site management infrastructure. In addition to tracking Oracle iStore sites, Oracle Web Analytics also by default tracks Oracle Partner Management sites set up using the Oracle iStore site infrastructure (Display Template management framework). In addition, Oracle Web Analytics integrates with other Oracle E-Business Suite applications for data capture and reporting purposes, including being able to report on Oracle Marketing campaigns.
Oracle Web Analytics tracks visitors who are registered users in the Oracle E-Business Suite, as well as those identified by the Oracle E-Business Suite as guest users. Authenticated users are tracked by their Oracle Trading Community Architecture (TCA) party. Anonymous visitors to both Oracle E-Business Suite and non-Oracle E-Business Suite sites are tracked using the Guest party in Oracle TCA. In the case where no party is found, the Oracle iStore Guest party is used. Setup of the Guest party would be an implementation step only in the case where Oracle Web Analytics is implemented for non-Oracle E-Business Suite sites; this setup is not required in the case where both Oracle E-Business Suite and non-Oracle E-Business Suite sites are present. A visitor to an Oracle E-Business Suite site will have all of the pages that he views assigned to a single unique visit, during his active navigation of a site. The Oracle E-Business Suite session will be correctly identified as authenticated or anonymous by the Oracle Web Analytics tracking engine. When a visitor ends his Oracle E-Business Suite session by signing out, Oracle Web Analytics also ends the user's visit. An Oracle E-Business Suite visit is assigned to a single party. A visitor may have multiple parties across visits. For example, Visitor 123 in Visit 1 uses Username A which is tied to Party X. In visit 2, Visitor 456 uses Username B which is tied to Party Y.
A visit to a site managed through Oracle Single Sign-On Server (SSO) is tracked in the same manner as an Oracle E-Business Suite visit. The visit will be correctly identified as authenticated or anonymous. When a visitor ends his SSO session, Oracle Web Analytics also ends his visit. An authenticated SSO visit may be assigned to a TCA party; if so, the user will be tracked by the party ID. When no such link exists, the SSO visitor is tracked by the Guest party in Oracle TCA.
Oracle Web Analytics supports the ability to track campaign visits created through Oracle Marketing Tracking solution, as well as scenarios where the customer directly links to a site from an external source.
A marketer creates a marketing campaign in Oracle Marketing and sends it to users via e-mail. A user views his e-mail in Yahoo and clicks on the link in the e-mail. The user is then taken to the Oracle Marketing tracking page, which records his response using Oracle Interaction History. The marketing page then forwards the user to the target page, in this case an Oracle iStore product details page. In this scenario, Oracle Web Analytics will capture the visit, along with the campaign schedule source code, and the referral source, in this case Yahoo. If the visitor should later in the visit create a shopping cart, the cart will be identified with the campaign source code tracked in the visit.
A marketing manager creates a campaign in Oracle Marketing and retrieves the campaign source code from the campaign URL. The marketing manager creates a Google campaign and supplies a link to their site with the embedded the Oracle Marketing campaign schedule source code ID (e.g., www.somewebsite.com/thegrid&msource=1234). Visitors who click this link from Google are taken to the target site. Oracle Web Analytics track these visitors and assigns their visit to the appropriate campaign schedule based on the campaign source code in the URL. Oracle Web Analytics tracks the first campaign associated with the visit even when a visit is associated with multiple campaigns, and tracks only Oracle Marketing campaign schedules that are supported by Oracle iStore.
The Oracle Web Analytics tracking engine allows the tracking of visits and page views for content that is cached by Oracle Application Server Web Cache.
Oracle Web Analytics allows Web managers to track non-Oracle E-Business Suite Web sites; these are known as third-party sites. Site designers can embed a JavaScript tracking tag into their site pages to track visits, visitors, page views, and site searches on their legacy Web properties. Static Web sites -- as well as dynamic Web sites (using Perl, JSP, ASP) -- can be tracked, as long as the JavaScript tracking tag is embedded in the site pages. For more information, see the chapter, Integration with Third-Party Sites.
Oracle Web Analytics supports the ability to track visits for pages using secure socket layer (SSL) connections (HTTPS protocol).
The sections that follow discuss visits and visit duration.
A visit automatically ends after a period of inactivity defined in minutes, as specified by the user in the Oracle Web Analytics Visit Inactivity Period parameter. A visitor who navigates to pages not tracked by Oracle Web Analytics would accumulate inactivity time. Thus, although the visitor may be actively navigating these pages that are not tracked by Oracle Web Analytics, the Oracle Web Analytics engine, after a period of time, considers the visit to have ended due to inactivity. If a visitor is required to login twice during a visit, the second successful login would automatically end the visit. The user would start a new visit after successfully authenticating himself. This would ensure that a visit is only tied to a single party.
Oracle Web Analytics tracks the duration of visits by adding the page view durations. When a visit ends (including visits that are automatically completed by Oracle Web Analytics due to inactivity), the duration of the last page view is excluded.
In partner scenarios, companies often have contractual agreements whereby they post advertisements or portal links on partner sites to generate traffic to their sites from the partner site. Each piece of counted traffic (whether a visit, a lead, or an order) is known as a referral. Administrators can create categories of referrals to group and track each referral. Using referrals is optional.
Oracle Web Analytics tracks the source of traffic to a page for the following type of analysis:
External sources of traffic: This refers to links from third-party sites used by a visitor to visit a page within a site. Oracle Web Analytics tracks the first external referral in a visit where a visit may contain multiple external referrals.
Which pages are used by visitors to begin such visits?
The exit traffic to external pages and links are not tracked.
Each external (non-Oracle iStore) site is associated with a single domain. Every page view captured for external sites must be first validated against the domains specified by the administrator in the Oracle Web Analytics Administration Application. The logic used is all page view URLs which match one of the following:
The domain name string matches exactly the domain of the page-view URL.
<Any sub-domain string> plus period (.) plus <domain name>
If a matching site is not found, an exception will be logged, and Oracle Web Analytics ignores the page view (the page view is not reported). If a site code is provided within a page view, the validation of the domain to the page view URL still occurs; an exception is logged if the validation fails.
The following example illustrates the behavior.
| Site | Domain | URL |
|---|---|---|
| A | Oracle.com | oracle.com |
| B | Peoplesoft.com | www.Peoplesoft.com Store.peoplesoft.com |
| C | Oracleusa | Oracleusa |
Note that site C contains a domain that is not in the correct format.
| PageView URL | Comments |
|---|---|
| oracle.com/product1.html | This would validate against Site A domain name |
| store.oracle.com/products.html | This would validate against Site A domain name |
| www.peoplesoft.com/products.html | This would validate against Site B domain name |
| www.oracleusa.com | This would not match any site, including Site C as the domain name specified for Site C does not contain the .com extension |
| Support.peoplesoft.com/index.htm | This page view would be validated against Site B; however, it would be rejected as it does not match any of Site B’s Site URLs. An exception would be logged. |
When a site has enabled the filtering of traffic, visits from IP addresses specified in the Oracle Web Analytics administration Traffic Filter box (in the Setup tab) are recorded by the tracking engine but not reported upon.
Oracle Web Analytics captures the following minimum data from the standard HTTP protocol information.
| Attribute | Description |
|---|---|
| Client Name | The IP address of the client |
| Referrer | Used by a client to specify, for the server's benefit, the address (URI) of the resource from which the Request-URI was obtained. |
| Cookie Parameters | The information required by the tracking engine to track visits. |
| User Agent information | The type and version of the browser and operating systems (e.g., Microsoft Internet Explorer Version 6, Apple Mac Operating System) |
| Attribute | Description |
|---|---|
| URI | The resource being requested from the Web server |
| Date and Time | The date and time for the request |
| Visit Identifier | Visit identifier to flag unique visits |
| Visitor Identifier | Identifier to flag unique visitors |
The code used in the client browser provides an extensible interface where Oracle Web Analytics administrators may implement and plug-in their own unique code to override the default attributes captured in the browser. For example, if the client code captures the page name from the JavaScript variables, the client code provides a mechanism to override this method and allow administrators to plug in their own code that would retrieve, for example, the page name from the page URL. The architecture allows client code that is provided by the administrator to be supported in future Oracle Web Analytics releases. The code will not affect the ability of Oracle Web Analytics development to modify the default behavior of the client tracking code. The following set of attributes are exposed as a Java API interface to allow for the extensibility:
Page Name
Page Code
Site Code
Context ID (i.e., Product ID or Section ID)
Oracle Web Analytics features an easy-to-use administration user interface, allowing managers to do the following:
Tracking Engine Setup: Configurable parameters include visitor and visit time-out definitions, as well as the definition of tracking filters to include or exclude traffic from certain sites.
Site Tracking Setup: Configurations allow for tracking both external and Oracle iStore customer sites
Report Dimensions Setup: Reporting dimensions can be set up for: pages, site areas, and Web referrals.
Recent privacy laws passed in the U.S. and the European Union (EU) have set precedents for the use and control of customer data. These laws require compliance by companies that operate within these countries, a movement which affects almost all Oracle customers. In the EU, the new standards, which target the wrongful collection of PII (Personally Identifiable Information), are stricter than those mandated in the U.S. To ensure that Web Analytics customers comply with privacy laws, Web Analytics provides comprehensive opt-in/opt-out mechanisms, with support for disclosure and access.
Tracking preferences can be managed in the Web Analytics administration user interface, in order to comply with the variety of global privacy management requirements. Included in the configurable parameters are the following:
Ability to define whether tracking cookies are persistent or non-persistent
Ability to define the duration during which anonymous visitors are tracked
Registered users as well as anonymous users can also opt-in or opt-out from Web tracking in the customer user interface
Oracle Web Analytics has built in security features which prevent malicious attacks on the tracking system. Page views are collected only for content delivered from web servers. Malicious visitors cannot record page views by simply calling the tracking code from their computers or by executing the client tracking code from pages saved to their computers. Visitors tampering with the client tracking code in order to manipulate the page view information are also minimized. For example, a malicious visitor cannot create new page objects in Oracle Web Analytics simply by removing the page code information and supplying new page names in the page views.
The Oracle Web Analytics tracking engine is compatible with various deployment options, for example:
Single Sign-On: Full support is provided for integrating with the Oracle Single Sign-On Server.
Web Cache: Web Analytics tracking can work along with Oracle Application Server Web Cache to allow companies to benefit from the performance improvements brought by Web Cache, without compromising Web tracking accuracy.
Oracle Web Analytics features a dashboard and the following reports, which are described in the Reporting chapter.
Visit Trend Report
Visitor Conversion Report
Web Referral Analysis Report
Web Product Interest Report
Web Customer Acquisition Report
Web Customer Activity Report
Web Customer Activity Trend Report
Web Campaign Analysis Report
Page Interest Report
Page Interest Trend Report