Scientists have relied on the Kepler scientific workflow system to help them automate experiments across many different fields using distributed computing platforms. An assigned director oversees the execution of a process in Kepler. Users must still choose the computing resources that will run the workflow's tasks. A workflow scheduler that can allocate workflow tasks to resources for execution is needed to further reduce the technical effort required by scientists. We evaluate numerous cloud workflow scheduling methods to determine what data must be exposed for a scheduler to successfully plan for the execution of a Kepler process in the cloud. We explain the value by discussing the advantages of each different kind of data about workflow jobs, cloud resources, and cloud service providers