Web Analytics Case Study

Client: Global Insurer with operating profit of over $5.5Bn

Objective: To deliver an advanced web analytics system across 5 European domains


  • Data capture must be highly detailed, every customer interaction must be captured
  • Reporting and dashboard portals must be of high quality
  • The data must integrate with other business data sources
  • There must be powerful analytics available for use against the data

Our route map to delivering these requirements was,

  1. Speak with client stakeholders to understand reporting and analytic requirements
  2. Offer invitation to web analytic vendors to pitch
  3. Compare vendor pitches to client requirements
  4. Run Proof of Concept with winning vendor
  5. Upon successful PoC, create production roll-out plan
  6. Gather and deliver detailed reporting requirements
  7. Gather and deliver detailed analytic requirements

The solution that was taken forward to PoC stage was SAS for Customer Experience Analytics (SAS CXA).

Here’s how it fared against the requirements, and why it was chosen,

  • Data capture must be highly detailed, every customer interaction must be captured

The SAS CXA solution collects “browser events”, which means that everything that happens in the end users’ browsers are captured.  This includes mouse movements, clicks, page loads, browser scrolls, as well as traffic sources, browser used, screen resolution, and the usual web analytics dimensions.  That pretty well covered that requirement!

  • Reporting and dashboard portals must be of high quality

The SAS CXA solution uses the SAS BI clients for dashboards and reporting.  These components offer high quality reporting features, user administration and report sharing features.

  • The data must integrate with other business data sources

The SAS CXA solution uses SAS Meta Data Server, SAS Data Integration Studio and SAS OLAP Cube Studio to build data warehouses and reporting structures.  These tools can integrate other data sources into the data model to make them freely available for reporting and analysis with the web data.

  • There must be powerful analytics available for use against the data

The SAS CXA solution uses the core analytics platform for forecasting, modelling and statistical analysis.

The Proof of Concept assessed system reliability, core data accuracy and reporting & analytic capabilities.

To read more about our criteria for Web Analytics capabilities, please read more about our web analytics consultancy services.


Reporting and Dashboards:

SAS BI Example

Reporting and BI Dashboards are delivered through an online portal environment, with central access and rights permission administration.  The first view of data an end user has is a series of Dashboards.  The objective of the dashboards is to highlight key data and trends on a single page, so that an overview of site performance, visitor volumes, conversion rates and revenues is immediately accessible. Dashboards highlight trends and snapshots of key data such as bounce rate, volumes, conversion, page views, average time on site, and so on, using a wide range of visualisations such as traffic lights, sliders, bar and line graphs and tables with conditional highlighting.

Once the key trends or alerts (notifications triggered by KPIs having exceeded a set limit, or fallen outside of a range) have been identified in the dashboard, the users then drill down into detailed reporting.  Detailed reports are powered by SAS Web Report Studio, and built upon OLAP cubes and tables in the data warehouse, and enable drill through, slice and dice and advanced graphical representations of data.

A suite of reports were delivered to fulfill the requirements of all client stake holders.  These included,

Marketing Views,

  • Source of business
  • Paid and natural keyword performance
  • Referrer and affiliate reports
  • Email marketing performance
  • Etc…

Operational Views,

  • Page load times
  • Slow loading content
  • Javascript errors
  • Broken links,
  • Etc…

Commercial views,

  • Goals (sales, quotes, etc…)
  • Goal conversion rate
  • Premiums
  • Cross-sell
  • Etc…

Data Integration

A key requirement of the system was that it should be able to integrate the data collected by the CXA product and the client’s back end data base.

SAS DI Example

This was achieved by populating the “transaction complete” page with the customer and policy ID for that transaction.  That was then collected into the web data model, and could be used as a join key to the back end systems.

Using SAS Data Integration Studio, it was then possible to integrate the data sources into a single data model and create reports that would not be possible without such data integration.  For example, it was possible to create reports on the customer renewal rates for each source of business (Search engine, affiliate, referrer, etc…).  This would be impossible if the data were not integrated because the renewals process is entirely offline.

Advanced Analytics

SAS Enterprise Guide ExampleWith an advanced web analytics system comes lots of data (that’s the point!).  In order to deliver maximum value from the data, we need to deliver advanced analytic capabilities such as forecasting, modelling and statistical analysis.  For example, using statistical measures and forecasting techniques it is possible to define control limits for a measure, such as bounce rate, and to create an alert when this measure exceeds this range in a statistically significant way.  It is also possible to optimise the performance of your website and traffic drivers with goal seeking techniques.  Statistical models will model the sensitivity of the target variable to changes in website drivers (click rates, page load times, PPC traffic, etc…).  By doing this, it is possible to conduct a cost/benefit analysis on operational drivers and to make the most efficient changes possible to drive the desired result.  For example, improving the page load times of the home page by 1 second could have the same effect on the bottom line as spending an extra £100 per month on Google AdWords.

To discuss your web analytics requirements, whether a quick win solution or advanced solution such as SAS CXA, please contact us.

Add Comment Register

Leave a Reply