Web Analytics: Failure to prepare is preparing to fail

John Wooden preached to his athletes, who were recently honored as being the greatest coaches in American sports: “Failing preparation is preparing for failure.”

Wooden’s lessons are a great way to achieve success on the court and in your life. This classic wisdom also applies to those who want to succeed in Web Analytics.

Web analytics can help companies capture valuable information about user behavior, gain valuable insights, and have a distinct advantage over their competition.

Web analytics initiatives often fail to deliver the insights that businesses want. Many companies are eager to get immediate value from newly implemented web Measurement tools, but they ignore the complexity of Web Analytics, which requires significant preparation, including preplanning.

If they are not properly prepared, web analytics initiatives will be hindered by data overload, additional time spent authenticating data, excessive reportage, and other issues.

Four key steps are outlined below for a proper Web-analytics setup.

Approach web analytics not as a report solution but as a solution for insight and optimization

Many online businesses take the easy way out when it comes to Web analytics. Many online businesses focus on collecting data and producing reports rather than making useful insights about the data most valuable to their company.

The challenge is not to gather raw data but rather to analyze it and follow a disciplined process of improvement and optimization.

For businesses to get the most out of their Web analytics efforts, they must first break down data and analyze every element.

In order to better understand specific behavior, qualitative data should be collected and analyzed via user surveys and usability testing.

The quantitative and qualitative data can be used to make recommendations for improvements. These are then tested using optimization techniques, such as multivariate and A/B testing. These tests will help guide the company in its improvement decisions.

Organizations that focus solely on data reporting cannot evolve to an insight-and-recommendation focus overnight. For data analysis and optimization to be successful, both technical and human capabilities are needed. These should be considered and developed at the beginning of Web analytics planning.

Develop a web-measurement strategy.

Businesses need to develop a Web-measurement strategy during the initial planning phase of Web analytics. The tactical plan outlines what the selected Web-analytics tools should measure, as well as how data collection, Web analytics reports, and data analyses will be aligned to the company’s strategic initiatives and overall business requirements.

Web measurement strategies begin with the definition of goals for the website of a company. This will determine the key performance indicators that need to be tracked. The last step is to identify supporting metrics and user behavior that impact the KPIs.

This three-step approach ensures that only the most important data is collected and analyzed.

Imagine a website whose primary goal is to encourage visitors to sign up for a free trial and then buy its Web service. Web measurement strategies would identify the conversion points for buy-and-try actions that should be tracked and measured. The KPIs will also include supporting metrics to help answer specific questions. For example, where do users go from just being site visitors to trial users and then buyers? How long does it take from the trial to the purchase? And how do these behaviors vary between different user segments? What are the most popular actions and content that trial users take, and how does this vary by user segment?

Create a Web analytics implementation plan.

The amount of data that Web analytics tools can collect is almost infinite. Too much data can be problematic, especially in the early stages.

  • The flood of data can cause the analytics team at a company to waste time separating the useful data from the unimportant data.
  • The analytics team must also verify the accuracy of the data and meet the numerous requests for reporting from different parts of the company.
  • Lastly, these distractions can prevent the analytics team from properly engaging in analysis and optimization efforts, which are the real payoffs for a Web Analytics project.

To avoid data overload and to reduce the amount of data, companies should implement Web analytics in phases. In the first phase, the company would measure only the behaviors that are most important to them. The web measurement plan would determine this. The Web Analytics Team should then ensure the accuracy of the data, analyze it, and make recommendations for improvements. They will test these recommendations before moving on to the second phase. A phased approach to implementation will provide immediate value for the business and have a positive impact on its success over time.

Be prepared that web analytics data may not be 100% accurate

Validation of data accuracy is required after a web analytics tool has been implemented and data has been collected and reported. Validation includes quality assurance testing (QA) in a development environment and comparing the data to the company’s existing measurement tools or external data sources.

Validation should be quick and confirm that data is within a tolerance for accuracy. Depending on its importance to the business, data that does not meet the tolerance should either be rejected or reevaluated.

The Web Analytics Team should begin using validated data as soon as the process is complete by performing analysis and optimizing efforts.

Unfortunately, many companies—particularly those new to Web analytics—spend too much time on the data validation stage. This is because most companies do not realize that their data may never be 100 percent accurate or consistent across data sources due to external factors.

JavaScript, for example, is not enabled in all Web browsers. Web users may also reject or delete cookies that are vital to identifying visitors. Data that is compared to a company’s existing measurement systems or third-party sources of data will always be inconsistent because measurement methods are fundamentally different.

A Web analytics department will spend an excessive amount of time validating data to meet the expectation of accurate data. This can cause the team to delay critical data analysis and optimization.

When the Web analytics report does not meet the expectations for precision, senior management is likely to distrust it, dismiss the data analysis, and lose faith in the Web analytics project.

The following principles can help avoid problems during the data validation stage if they are implemented in the planning phase for a Web analytics initiative:

  • A level of tolerance acceptable for data accuracy
  • Understanding that Web analytics data is not precise but within acceptable tolerance levels
  • Data collection best practices that minimize inaccuracy
  • The process of data validation and QA for Web analytics is concise and reduces the risk associated with excessive data validation.

Leave a Reply