Research timelines slip for a structural reason, not a performance one. Project managers estimate analyst effort well, because effort is something the organization controls. They estimate data availability poorly, because availability depends on systems outside the project boundary that have grown steadily more restrictive.