Monday, March 30, 2009


If someone is worth recommending...

...then it’s worth writing a good recommendation.

And here I don’t just mean a positive recommendation, though that, too, of course. I mean that you should spend some time writing it well. And I mean that you should check it for spelling, punctuation, grammar, and style.

For a case in point, look at this actual recommendation I got last year, for a graduate student we were considering:

[Name] is viewed in the technical and executive community as some one, who using her teachnical ability, and ability to understand the root cause of seemingly dirrerent positions teken by various inter related groups, is able synthesize a solution that enhances the design and makes various party comfortable.
That’s a copy-and-paste job, with the candidate’s name removed; all errors were in the original.

As it happens, and as you might guess, the author of that recommendation is not a native English speaker. Even so,

  1. his spoken English is not that bad, so this is just carelessness, and
  2. he could have — should have — had a native speaker check it.

So let’s correct what’s written, and have another look:

[Name] is viewed in the technical and executive communities as someone who, using her technical ability and her ability to understand the root cause of seemingly different positions taken by various interrelated groups, is able to synthesize a solution that enhances the design and makes various parties comfortable.

OK, now that the writing is tolerable... what does it say?

I certainly haven’t a clue. And that’s it, in its entirety.

A recommendation is meant to tell the reader — me, in this case — why he would want to give the candidate a position. What does she bring to my organization, my program, my project? What specific skills does she have that would benefit me? What do you know about her that you can tell me, that will make me jump out of my seat? Tell me one or two things that she’s succeeded in that are relevant to me.

But the example here is completely generic. It says nothing about what the writer knows about the candidate (just some vague thing about how she’s viewed in the community). It says nothing about what she’s done. It says nothing about how her skills match what I’m looking for. She has technical ability? In what areas? What has she accomplished with it?

It describes, in a bizarrely complex clause, some knack she has for understanding people’s ideas... but, again, it’s vague about it, and doesn’t say more.

And, in the end, what’s the only thing the author tells me she’s done? It says that she can “enhance the design” (of what?) and “make various parties comfortable.” What parties would that be? And comfortable? That’s like when your parents used to say that the friend of a friend whom your parents were trying to fix you up with was “very nice.”

What was I supposed to do with that recommendation? What I did was discard it. And then I sent email to some contacts of my own, who could give me some real recommendations about that candidate.

Not everyone is going to go through the trouble. If you’re applying for something and most of your recommendations are like that, your application will most likely go into the trash.

If you’re writing a recommendation, take the time to write one that will truly reflect well on the applicant, and that will be useful to the people making the evaluation. Say specific things, and make the personal connection — how you know the applicant and what you have seen that impresses you. Why would you hire her? Tell me that.

And if you wouldn’t hire her — if you’re writing a recommendation pro forma because you didn’t want to hurt anyone’s feelings by saying no — then, by &deity, don’t write it! Say no, please. Because a recommendation like this is worse than none at all, so in the process of not hurting her feelings, you’re hurting her chances.

If you’re soliciting recommendations, you’re in a tougher situation. You need to pick three or four people you can trust to do it right. Use your sense in making that selection. Maybe you can find a way to see some recommendations they’ve written in the past. Maybe you just have a gut feeling that they’d write good ones. Choose carefully. Then discuss with them what you want them to highlight, and what you think you’d like them to say.

They need to write it in their own words, and it has to be an honest recommendation, but that doesn’t mean you can’t help them craft it. Remind them of your most relevant skills. Remind them of the times you’ve worked together successfully, and point out the key elements of that success, and the value of it. Then let them go off and write.

And remember that each recommendation needs to be custom-written for the position you’re looking for. You may want different people writing references, depending upon where you’re applying. Does this or that professor have a close relationship with the school or company you’re hoping to go to? Pick the one who does, and take advantage of those connections. A recommendation means more if the reader knows and trusts the person making it. Don’t just use the same people and don’t try to recycle their recommendations. Even if this position may be similar to another, it probably won’t be exactly the same.

Finally, if your thesis advisor or previous employer isn’t one of those you’re getting a recommendation from, be prepared to explain why, in a positive way (don’t say why you didn’t get along, but say why the recommendations you did get are more relevant). The question might not come up... but if it does, you’ll need a good answer at the ready.


Marcy Webb said...

Hi, Barry.

Thank you for your post.

What you highlight in your well-articulated points is that some people do not know how to write very well, and/or wait until the last minute to compose the recommendation. If every person charged with writing a rec. and with soliciting recs. would read your post, there would be more successful job search campaigns.

The Ridger, FCD said...

You exclude the possibility that I don't want her to get the job. That a badly written (in this sense) recommendation is really a massive Gricean violation with intent...

But surely people wouldn't do that, would they?