Adversary emulation has become an increasingly common type of engagement where red teams look to known threat groups to inspire the actions and behaviors used. While scoping activity might make operating easier, emulation introduces a new set of challenges to planning. How do you know how an adversary behaves? What do you do if you only know part of the picture? How do you turn all of that into a plan? In this talk I’ll examine how we can start building an adversary profile from the open source intel in MITRE ATT&CK. Open source intel often doesn’t give a complete picture of an adversary, and I’ll talk about some of where these gaps come from, how to spot them, and some ways of filling them in. I’ll work through a process for turning the profile we’ve created into an adversary emulation plan expressed in ATT&CK and how we can stay aligned with that plan as we operate.