How do you determine which user research methodologies to use when timelines are tight and data is limited?