Determining which indexes are going to be of use is primarily done by understanding the queries that are going to be run against the data. You can't do one without the other. But, the most important decision you have to make is where to place the clustered index. Take your time to understand both your storage needs and your query needs. Generally, the best place to put the clustered index is on the most common access path to the data. This can be the primary key, but it won't necessarily be so.