Most companies have a web analytics solution implemented, but a small minority is using it to its full potential. After listening to Geary’s Nicole Rawski present at Ad:Tech last week, it’s clear that segmenting is an area where marketers are missing the chance to drive their online ROI .

Here are answers to the top concerns about segmentation.

How can I prove value and ROI from segmentation?

ROI correlates to your online business goals and metrics. Use segmentation to refine your online marketing initiatives and directly relate them to your corporate business objectives. For example, “if we increase conversions by .03%, we’ll increase revenue by XX%. Statistics like this are hard to ignore and translate data into revenue and goals.

How much budget should I allocate for segmenting and testing?

Segmenting should be a regular part of analysis, and testing is integral in optimizing web performance. Budget should reflect the level of intelligence and reporting needed to reach desired ROI levels. If budget is a concern, start with free tools to prove the value of analytics tracking and testing. This will help you build a case for more sophisticated tools that can improve online performance at a much greater level.

How big should my sample size be?

Sample sizes need to be large enough to produce statistically significant data. Aim for a 95 percent confidence level, and use available free calculators to determine your particular sample size based on your testing metrics.

What is the ideal timeframe for testing?

Timeframes vary, but typically, a testing period should last at least a week. This is independent of traffic volume or sample size as visitor trends change throughout a week and weekend. If your product or service has other seasonality to consider, take this into account when creating your testing parameters. For instance, December is the worst time to test online consumer behavior as patterns are irregular and likely to sway data.

On Twitter? Follow Robyn at @RobynFreye. Follow Geary at @gearyi.