There is no shortage of research on the importance of good teaching. For decades, study after study has shown that there are large differences in effectiveness from one teacher to another and that these differences can have a lifelong impact on students. A recent study that tracked 2.5 million students over 20 years determined that those with highly effective teachers "are more likely to attend college, earn higher salaries, live in better neighborhoods, and save more for retirement. They are also less likely to have children as teenagers." Yet there has been little research on exactly how schools can get an accurate picture of their teachers' performance in the classroom. States and school districts have been left largely to their own devices when it comes to this singularly important task. The results have been disastrous. As documented in 2009 study "The Widget Effect: Our National Failure to Acknowledge and Act on Differences in Teacher Effectiveness," most teachers are evaluated infrequently and according to low standards. They rarely receive feedback that helps them improve. Nearly every teacher is labeled "good" or "great," no matter how much progress their students are making. In the end, the entire profession has suffered from this negligent approach. Groundbreaking new findings from the Bill and Melinda Gates Foundation's Measures of Effective Teaching (MET) project hold the potential to answer crucial questions about how to assess teachers' performance. For the past two years, MET researchers have conducted a research project of unprecedented scope, involving 3,000 teachers in six school districts across the country. Using gold standard research methods, they have tested a number of evaluation approaches, including student achievement data, classroom observations, and surveys of students. Their most recent report, "Gathering Feedback for Teaching," provides a wealth of practical implications for improving teacher evaluations. This paper is intended for policymakers who are developing better teacher evaluations and are looking for ways to apply new research findings quickly. It summarizes the lessons from MET and provides recommendations on how these lessons can be applied right now. (Contains 16 footnotes.)