In one of my recent posts, I talked about the importance of getting feedback on your work. But feedback is a tricky beast. If it's too light, it might not challenge someone to do better. Too harsh, meanwhile, and it can kill someone's drive to even bother with something.
So, how do you reach the middle ground?
Own Your Opinions
This is probably one of the best pieces of advice I've ever picked up. It doesn't just apply to feedback, but all aspects of communication: Own your opinions. What does that mean?
See if you can spot the difference:
Celery tastes terrible.
I think celery tastes terrible.
One of those is an opinion, and the other one is an opinion stated as a fact. If I walk up to a friend and say, "Celery tastes terrible," and they happen to like celery, I've immediately put them on the defensive. If they don't want to be 'wrong,' they have to come out against me, and I've created a conflict where one doesn't have to exist. Meanwhile, if I say, "I think celery tastes terrible," then my friend subconsciously knows it's an opinion and isn't compelled to defend his or her taste.
It's easy to apply this to feedback. If you're critiquing someone's style or content, and assuming you want to stay on good terms with them, make sure to let them know how you're feeling about it. Obviously, factual errors need factual statements, but if you think an article would flow better if it was arranged a different way, for example, make sure you tell them, "I think you could switch these two paragraphs around, and it would read a lot better."
That way, you're collaborating, not dictating.
Point Out the Good With the Bad
Positive reinforcement trumps negative reinforcement every time. If all you do is point out what someone does wrong, they're going to associate that feeling with you. Eventually, unless there is a remarkable level of trust and camaraderie built up, you're going to stop being asked for feedback, period.
Because nothing (or at least very little) is universally bad.
If you like something about a piece of work, say so.
"You did a great job cleaning up that code."
"I liked how you used repetition to get your point across."
"That was a fantastic ham sandwich."
Those are three very basic examples, but you get the idea.
Speaking of sandwiches, this is not an endorsement of the "Praise Sandwich," where you alternate between giving someone praise, followed by constructive criticism, followed by praise. (Family Guy fans will recall Stewie doing this to Brian at one point.) That's a ludicrous concept because it creates an unreal expectation of praise. What happens when there's more that needs fixing than the stuff that works well?
You're either going to cut your criticism short, or you're going to fish around for some barely passable piece of praise — and neither of those is desirable.
Be Honest, Be Specific, and Actually Offer Advice
If you follow the steps above, the last one is easy: be honest. Don't pull your punches for the sake of being polite. If something is wrong or strikes you as needing improvement, point it out. Constructive feedback will make so-so work good and good work great.
That being said, how you point it out matters.
Like I said above, foster an attitude of collaboration by owning your opinions. If you think something needs work, say that: you think it needs work.
Even better, be as specific as possible. Never, under any circumstances, say, "I'd do something to fix this." That's useless. You undermine your own feedback when you give vague, general advice like that. After all, what's the response to that? "Gee, thanks. I guess I'll do something to fix it then."
Instead, offer a sample of what would make something work for you.
It's the difference between:
Paragraph two needs work.
I don't think you're being very clear in paragraph two. It gets a little confusing when you talk about containers and VMs.
Paragraph two starts strong, but I think you lose the message partway through. I'd separate the talk about VMs into another paragraph and just make that paragraph only about containers.
The third time's the charm there. You've not only owned your opinion, but you've also pointed out what you like about something before moving onto the ways to improve. Subconsciously, this is huge. You've validated the work while offering a collaborative way to improve it.
Combine all three, and you portray yourself as a helpful, knowledgeable, approachable person who can help improve work without relying solely on negativity.