How to do a usability test

First, when you are conducting a usability test, it is important to understand exactly what data you should be collecting. You should not run a test without first deciding on what data is required to address your business challenges. Plan ahead! Second, in a usability test, you don't just watch users. You must collect data that reflects how customers actually use your products and services. This is easier said than done.

One of the core ideas of usability is that you need to measure what people are actually doing. Sure, you need to watch them, but that is not enough. You have to understand what to record as they are using your web site, product, or service. 

Please remember that you aren't really testing people. A usability test is not meant to test intelligence! Instead, you are testing your products and services. You are testing how people interact with your web site. Beware, if you start thinking that you are testing people, you will go down the path of blame. Before you know it, you will blame users for problems with your technology. You will start to think that customers are stupid, and that is something that goes completely against usability. The failures you see and record are the failures of your web site.

Below is a list of metrics that are based on software usability research.

Usability Metrics

Effectiveness

• Percent of tasks completed

• Ratio of successes to failures

• Workload

• Number of features or commands used

Efficiency

• Time to complete a task

• Time to learn

• Time spent on errors

• Percent or number of errors

• Frequency of help or documentation use

• Number of repetition or failed commands

User Satisfaction

• Rating scale for usefulness of the product or service

• Rating scale for satisfaction with functions and features

• Number of times user expresses frustration or anger

• Rating scale for user versus technological control of task

• Perception that the technology supports tasks as needed by the user