Author: Michael Senger CEO, StoneMass

Part 2 – Where do I begin to start analyzing web data?

Knowing that you have so much data available to you can be both enlightening and daunting. The question you need to ask is where I begin. Start with your company goals for the year. If you don’t have any, well that’s a blog topic for another day. From your company goals, you can create specific marketing objectives that transcend to your website. From these marketing objectives, you will identify and measure key performance indicators (KPIs). Your KPIs are key web activities or actions that you want to focus on.

It is unrealistic to state “I want to merely increase web traffic to your website”. It is more useful, for example, to define a goal of increasing web traffic by 10% to support a product launch for September.

Let’s consider the following company example:

Company Goal:
Acme Company sells consumer widgets and one of their 2009 business goals is to improve customer loyalty.

Web Marketing Objectives:
Benchmark, measure and improve online customer satisfaction KPIs for the 1st quarter of 2009.

Key Performance Indicators (KPIs)

Example:
Retention KPI = Ratio of “Returning visitors”/”All visitors”
What it measures: determines how you are doing at retaining visitors
Metric: increase by 10%
How: promote webinar and product announcements

Action:
First determine your KPI benchmark to identify a % goal increase. I usually lean towards measuring the same period before. Since we are measuring performance for the 1st qtr of 2009, I would recommend you compare metrics to December Qtr 2008. Use your judgment here; if your company has strong cyclical sales you may want to measure the same period from the year before.

Next: Part3 What do my KPI measurements mean?

Author: Michael Senger CEO, StoneMass

Why is there a certain ‘voodoo’ associated with web metrics and analysis? I would think most business owners and marketers when asked “do you need better analysis tools to measure your campaign effectiveness?” would immediately raise their hands. We all want better measurement of our $’s performance but are averse to understanding the data mechanics or even how to go about generating an insightful report.

This ‘Far Side’ comic reminds me of a company I previously worked for. This was the typical response from fellow marketers when asked what kind of metric reporting is being done to measure campaigns. Simply, no one wanted to own the metrics data or be held responsible for reporting it.

The Wikipedia explanation of the phrase origin “Lies, damned lies, and statistics” states the following:

“…the use of statistics to bolster weak arguments, and the tendency of people to disparage statistics that do not support their positions.”

There lies our second issue; the numbers are accurate but the interpretation and usage by the individual can sometimes be suspect.

In this three part series, it is my intention to dispel this voodoo and to provide a methodology for evaluating and measuring the success of your online program .

Part 1: Keep it simple – Too much data is just that – too much data!

By 1969 in Vietnam, the kill ratio for American jets during dog fights was plummeting. At one time, 20 to 30 enemy planes were shot down for every one US jet. By the end of the decade the ratio was one to three. American jets and their technology were far superior to the Russian Mig Jets but a disturbing trend was happening.

When they interviewed the pilots that were highly successful in air to air combat, they found something in common. As fighter information technology advanced, so did the pilot’s dashboard and instruments. The successful pilots identified and focused on a few instruments; ignoring the rest. One successful pilot had even taped over redundant instruments. In the engagement of the air enemy, life or death decisions were based on key critical information.

Most web analytics software programs, Omniture, Google, etc. offer the user so MUCH information in their web marketing dashboard to send the user screaming down the hall in frustration. Keep it simple and too much data is just that – too much data.

Next : Part 2 – Where do I begin to start analyzing web data?