As investigators, we can't always get exactly to the evidence we want to prove. Sometimes it merely doesn't exist. Often, ethical and legal constraints keep us from being able to obtain the facts we definitively need to prove what we are investigating. It's easy to get lost searching for the unsearchable, pining for that one nugget that will help everything fall into place. But investigators don't have that luxury. So, we sometimes have to do what the computer scientists have done by pinpointing a font as a sign of trouble: We have to take a step back and look for clues elsewhere. We may not have direct evidence of wrongdoing, but we can scour the evidence in order to detect patterns that suggest wrongdoing. Alternatively, we can review the facts to see if we can find any that correlate with what it is we've been asked to help prove or disprove. This is not about making assumptions--we never say that because x exists, therefore y. Instead, it is about being able to look for solutions that advance our clients' knowledge, even if they fall short of the ideal solution.

We wrote in our pieces "What Greg Smith and Goldman Sachs Tell Us About Investigations" and "Hiring Due Diligence Should Include an Attitude Check" about how indispensable it is to talk to people during an investigation. No matter how thorough…