User research
Survey
Self-reporting information is extremely valuable. It is also the most efficient letting you reach out to thousands of users at once. However it lacks context so it is often difficult or impossible to discern the cause and effects if surveys are the sole data points. Following surveys up with interviews, or emails helps. Triangulating it with data from interviews and focus groups etc., also help strengthen the quality of the data by bringing some context via alternative complementary data.
Focus group
Efficient for getting feedback from many stakeholders at the same time. The problem is the quality of the feedback is at the mercy of the group dynamics which you have some but not complete control over. The information by itself is usually unreliable. In combination with some interviews, surveys etc though the outlier information can be caught and then these groups can be more valuable as supporting information.
Group task analysis
One method to get users to pragmatically talk about their work in a group setting is have them sketch their work flow together. It usually comes up with higher fidelity results than simple focus groups would or surveys.
Interviews
Less reliable than site visits, but second best way of gathering end user feedback is interviewing the users one on one. Without peer pressures associated with a focus group, end users are freer to discuss their working situation frankly and confidentially. To gain maximum benefit the user is often asked to bring in work samples and discussing them concretely. Interviews are the preferred method for software intended to address critical incidents. Users’ memory of these incidents are usually quite vivid, and the chances of catching a critical incident during a site visit is usually low.
User site visits
In order to assure the lowest risk in a product or service development, it is essential to understand the needs of the user with it being filtered by perceptions, even by the users themselves. Observing them in their work place gives the most reliable information on what their real work flow is, what their real information needs are, what their real experiences with a product is, etc. User site visits provide the lowest risk but also a moderately high cost, as one famous line goes, “It costs money because it saves money.”
Resources:
Baseline usability testing
For existing products it is a lot easier to tell if a new design is an actual improvement instead of a simple channel change: first conduct a baseline usability test to set the levels where the new design must be a significant improvement on the current design.