John Wooden, who was recently honored as the greatest coach in American sports, would often preach to his players, "Failing to prepare is preparing to fail."

Wooden's lesson is a key to success both on the basketball court and in life—and it is classic wisdom that also holds true for those seeking success in the field of Web analytics.

With proper preparation, Web analytics will allow companies to capture rich user-behavior information, derive valuable insights from that data, and gain a clear advantage over their competitors.

But far too often, Web-analytics initiatives fail to provide the insight that most businesses desire. Companies are so eager to obtain immediate value from their newly implemented Web-measurement systems that they overlook the complexity of Web analytics, which requires a significant investment in preparation activities, including pre-planning.

In the absence of effective preparation, Web-analytics initiatives are likely to be hampered by data overload, extra time spent authenticating data, excessive reporting, and other issues.

Detailed below are four key steps recommended for proper Web-analytics preparation.

Step 1: Approach web analytics as an insight and optimization solution, not as a reporting solution

Many online businesses choose the easy route to Web analytics. They focus their efforts on collecting data and generating reports rather than producing usable insights on the data that are most valuable to their business.

That's because the challenge for most businesses is not in gathering raw data but in analyzing that data and successfully following a disciplined improvement and optimization process.

To capture significant value from Web-analytics efforts, businesses must first drill down, break their data into smaller parts, and analyze each distinctive element.

To better understand specific behaviors, the data-analysis stage should be followed by qualitative data collection and analysis via user surveys, usability testing, and other qualitative research methods.

Recommendations for improvement can be made from the quantitative and qualitative results, which are then tested through optimization efforts, such as A/B and multivariate testing. The results of those tests will guide the company's improvement decisions.

Organizations that focus solely on data reporting cannot evolve to an insight-and-recommendation focus overnight. Both technical capabilities and human expertise are required for successful data-analysis and optimization efforts and should be accounted for and developed during the initial Web-analytics planning stage.

Step 2: Develop a Web-measurement strategy

During the initial Web-analytics planning stage, businesses need to establish a Web measurement strategy. The tactical plan details what the chosen Web-analytics tool should measure and outlines how data collection, Web-analytics reports, and data analysis will be aligned with the company's goals, strategic initiatives, and overall business needs.

The development of the Web-measurement strategy begins with defining the goals of the company's website; doing so determines the specific measures for success, or key performance indicators (KPIs), that should be tracked. The final step involves identifying supporting metrics and user behaviors that have an impact on the KPIs.

The three-step process ensures that the most critical data, which is directly linked to the business's needs, is collected and analyzed.

Consider the process for a particular website the primary goal of which is to motivate site visitors to sign up for a free trial of its Web service and then purchase it. The Web-measurement strategy would define the buy-and-try actions as KPI conversion points that should be measured and tracked. Various supporting metrics that influence the KPIs would also be defined and used to answer specific questions, such as these: Where do users fall off the path from being just site visitors to becoming trial users and, ultimately, buyers? What is the duration of time from trial to purchase, and how does that behavior vary among different user segments? What is the most popular content consumed, and what are the most common actions that trial users and purchasers take?

Step 3: Develop a phased Web-analytics implementation approach

Web-analytics tools deliver an almost limitless amount of data. However, too much data, particularly in the early stages of a Web-analytics initiative, can be problematic—for several reasons:

• First, the flood of data will cause a company's analytics team to waste valuable time distinguishing the important data from the less useful data.
• Second, the analytics team will need to verify the data's accuracy and accommodate numerous reporting requests from various parts of the organization.
• Finally, such distractions will delay the analytics team from engaging in proper analysis and optimization efforts, which are the true payoffs of a Web-analytics initiative.

To avoid data overload, businesses should divide their Web-analytics implementation into phases. The first phase would involve measuring only the highest-priority behaviors, which would be specified in the company's Web-measurement strategy. Once the data is captured, the Web-analytics team should ensure its accuracy, analyze it, deliver recommendations for improvement, test those recommendations, and then move on to the next implementation phase. A phased implementation approach will deliver immediate value to the business and have a significant positive impact on the business's success over time.

Step 4: Be prepared for Web-analytics data not to be 100% accurate

Once a Web-analytics tool is implemented and data is collected and reported, the  data's accuracy must be validated. The validation stage includes quality assurance (QA) testing in a development environment and comparison of the data with the company's current measurement tools or third-party data sources.

The validation stage should be brief, confirming that the data is within an accepted tolerance level for data accuracy. Any data that is not within the tolerance level should be either rejected or re-evaluated based on its degree of importance to the business.

Once the process is completed, the Web-analytics team should immediately begin using the validated data by implementing analysis and optimization efforts.

Unfortunately, many companies—particularly those that are new to Web analytics—spend too much time on the data-validation stage. That is largely because many companies don't realize that due to factors that are beyond their control their data will never be 100% accurate or consistent with other data sources.

For example, JavaScript is not enabled on all Web users' browsers, and Web users can reject or delete cookies, which are essential for identifying site visitors. Furthermore, data that is compared with a company's current measurement systems or to third-party data sources will invariably be inconsistent due to fundamental differences in measurement approaches.

The expectation of precise data will cause a company's Web-analytics team to devote an excessive and unnecessary amount of time on data validation, which can delay the critical data analysis and optimization efforts.

More important, when the expectation for precise data is not met, senior management may distrust the Web-analytics reports, discredit the data analysis, and potentially lose confidence in the entire Web-analytics initiative.

Problems associated with the data-validation stage can be avoided if the following principles are established during the planning phase of a Web-analytics initiative:

• An acceptable tolerance level for data inaccuracy
• An understanding throughout the organization that Web-analytics data will not be precise but will be within the accepted tolerance level
• Best-practices in data collection that minimize data inaccuracies
• A concise Web-analytics QA and data-validation process that mitigates the risk of excessive data validation

* * *

On the first day of practice for a new basketball season, John Wooden would tell his hotshot recruits, "Gentleman, today we're going to figure out how to put our shoes and socks on." The purpose of Wooden's seemingly mundane task was not just to help the players avoid blisters, but to instill in his team the mentality of doing everything possible to prepare for success.

Companies that are embarking on Web-analytics initiatives should follow Wooden's fundamental lessons of preparation and begin by following the four steps detailed above.

If those steps are carefully considered and well executed, businesses can greatly increase their chances of extracting maximum value from their Web-analytics efforts and achieving Web-analytics success.