Integrated analytics: buyer beware

You would think integrating your analytics software with content management, eMail marketing, eCommerce, and other web applications would be a simple and straightforward task. But there is significant miscommunication in the marketplace about what constitutes true integrated analytics.

So how do you shop for and implement a fully integrated analytics package into your website? Today there are three basic choices available when selecting analytics tools for measuring web metrics: stand-alone analytics tools; connected tools; and integrated analytics:

»» Stand-alone analytics. Stand-alone tools are analytics packages that are not integrated with the CMS or other website management applications. Examples include SPSS Web Analytics, SAS Web Analytics, and WebTrends Analytics.

With stand-alone analytics tools, you run the application on demand. Data is fed automatically from your website into the tool. The tool processes the data, calculates the key metrics, and generates reports showing how well or poorly your site is doing.

Metrics reports are read on the screen or offline. After reading and studying the analytics reports, you and your team can discuss ideas for improving key performance indicators (KPIs).

Changes are then implemented in web copy and design using your CMS, or in the case of static websites, by hard-coding through an IT team. After making changes, you can measure the performance of the revised pages with the analytics tool and learn whether your changes actually improved KPIs.

There are a number of drawbacks to using an analytics package that is not integrated with your other website management applications, especially your CMS.

To begin with, the analytics package and the other website management applications can't directly communicate with each other. In addition, metrics reports produced by stand-alone analytics packages can't access and do not contain page-level detail maintained by the CMS such as information about the page author, publishing specifics, duplicate page use, and hierarchical layout.

Stand-alone analytics packages don't give analysts a quick or easy way to change or remove content based on site analysis. If an underperforming page is identified within a stand-alone analytics tool, it may be difficult to track down the page author to request changes.

»» Connected analytics. You can connect a stand-alone analytics package to your CMS and other applications by having a programmer write "hooks" or "page tags". Examples of this category of analytics tools include Google Analytics and Omniture.

Page tags are short segments of code, usually in JavaScript, that pass information about the page and visitor to the analytics package. Page tags can loosely integrate certain analytics packages to your content management system and other website applications (e.g., shopping cart, email distribution, CRM).

Transmitting data to a non-integrated analytics tool via JavaScript introduces certain risks. If the code on the page does not execute due to a conflict with other JavaScript code, the page data may not be transmitted. Often times, these conflicts do not even manifest themselves in a browser error, making it difficult to track down problematic implementations and faulty data collection.

Even with custom programming and page tags, the ability of the analytics package to communicate with the CMS, eCommerce, and other web applications remains limited, since they were not designed to work together as a single, integrated system. In addition, any upgrade made to the CMS may cause the link to the analytics package to be severed, requiring page tags to be modified once again.

Analytics integrated within a CMS pose no such concerns as it captures the data on the server level. This also eliminates problems surrounding latency of JavaScript requests and the interference introduced when data needs to be collected on form submission or "on click."

»» Integrated analytics. The concept of building websites with directly integrated web analytics tools is gaining in popularity. This deeper integration, espoused by Bridgeline with our own iAPPS Product Suite, enhances communication between the analytics tool and the applications it monitors, so that knowledge gained through metrics measurement can be used in real-time to deliver the superior user experience best practices described in the Gartner report.