I’ve written letters for these three great postdocs.

“Kind to pets” is what a colleague calls positive letters of recommendation that don’t really say anything. Even worse is “word salad,” a series of highly predictable sentences that could be said about anyone.  I write tons of letters every year and read hundreds more. What makes a letter good? What gets your student or your colleague the job, the prize, or admission to grad school? What if you have to write for several people from your group for the same job? This is both harder and easier than you think.

It must be hard because I have just read hundreds of letters of reference for the 3 open positions we have and generally each one has only a paragraph or two of real substance. Here are some guidelines I hope are helpful.  Generally, think about what we want to know. Think…