The most important component I have found from working in a company with an open, expected and established culture of consistent feedback, is that objectivity comes from specificity. This leads to a much better outcome and people having less of an issue taking the feedback onboard as it removes emotions and concentrates on the issue and not the person.

My particular approach, learnt from a previous organisation I worked for, and which works brilliantly:

Take time to actually adequetely identify the 'specific' action/habit/issue that needs correction and then find a clear impact on a given stakeholder. I know this sounds obvious, but is often overlooked. A great way to find if you're are specific enough is that it should come down to something you can see or hear. You can't see someone "being too negative", only actions that lead to the subject being percieved so.

Now keeping the clearly identified action/s in mind, you can approach the person with an open mind, ask for permission to give the feedback (you want to set an environment conducive to the process), and then let them know the problem as well as how it is affecting customers/the team/you.

The impact not the negative action is possibly the crucial element here depending on the actual problem.

Of course, it is important to not come across condescending but it will pay off to not tell the person how to fix this behaviour, but let them offer it to you. I have found that many times people will not know what they were doing was wrong and come to the right conclusion themselves.

Work together to find a way to overcome the issue if necessary, and offer to help in the process or to point them in the right direction—for instance if persons problem comes down to time management, and one of your peers happens to have that as a strength, then maybe they could spend time together on a project and that peer could share the way they effectively manage their time.

After the exchange it's also important to make sure that they know everything is 'cool' between you two, so later in the day/week it may be advantageous to strike up conversation on a completely unrelated topic, or just ask how their week is going, even invite them for a drink after work etc.

If done correctly this approach will work literally every single time. It did not fail me once in four years.

(I'm all for being polite and framing negative feedback positively, I try it myself, but ultimately it's all completely pointless. If someone can't take "I don't like that you did x, it should be why because z" without it starting with "I really love this and I love what you're trying to achieve ... but I guess you made that choice for a reason, so keep up the good work!" they might need to rethink creating things... someone shouldn't need to say they care and say they respect the person: if someone is giving you feedback then it's already clear they care... but then I guess that's the human brain, we're weird creatures that constantly need positive reinforcement)

But to pad everything with sucking up to a total stranger? Nah. At least not when talking about general or bigger stuff, as opposed to personal suggestions or projects. I show people I care; just not necessarily about them. They kinda have to earn that, first.

I don't, I see sucking up as sucking up. E.g. If someone says all black people should be killed, do I even have a right to assume they have great, superb, excellent intentions? Should I really start my response with lauding those? Nah. There's limits.

"I can see you are trying to appeal to teenagers, but the copy comes across as very young."

If I just said, your copy is childish, the recipient would likely think "doesn't he realise I'm trying to appeal to teens? God did he even read the site?"

Without context it's harder to get people to understand and act on your concerns, as most peoples first response is to already be on the defensive and try to find reasons not to implement your feedback.

But all that assumes good intentions behind what you criticize. They're all happy feel-good "we're in the same boat here" examples.

How would you criticize extreme greed or callousness? Or blind obedience to state authority, and hand-waving away murder? I've had bitter fights both with someone who kinda glorifed the RAF (Rote Armee Fraktion, not Royal Air Force), as with about anyone I run into on the web who says things like "Assange should be shot on sight, and without trial, traitor blah blah". And then there was this dude on a unmoderated forum who kept mocking someone else for being lesbian and having been raped as a child.

Reacting politely to things like that would have made me feel kinda dirty, you know? Beyond certain points I simply give up trying to change someone's opinion, and try to make it costly for them to have it. There are billions of people, some are complete sociopaths, or are deceived by the same; the more energy you waste on them the less you have for the rest. And being friendly to something disgusting drains me. Attacking it nourishes me. So that's that.

But it's important to understand the real point of such a technique- which is to prevent the other person from becoming defensive and shutting down. That is really what you're trying to avoid, and there are a variety of ways to do that, including but not limited to the above. Because if they are defensive then they aren't receptive, and you've failed in giving feedback.

As michaelochurch says below, framing the criticism as it applies to their goals (as opposed to yours) is one way to make their ears perk up and keep listening.

Always be kind and positive about what you say though. Assume that their intentions are, like you, to create a good product or do the right thing.

Another direction that you can take this is in getting other people to agree with you and speaking indirectly through all of them. However this is complicated to do right...

The advantage of finishing on a high is it often lets the conversation continue so that your subject can ask for clarification if they need it.

I think you just have to contextualise the feedback to show you understand the issue then explain where you see the problem. Then if asked give your advice on what you would do to fix the issue. Leading with your chosen solution is another common mistake. There may be good reasons they didn't do that to start with, and the problem you were addressing becomes lost.

People don't want to listen to you if you're sniping at them for how they failed to serve you, but if you can indicate to someone a way they could more effectively pursue their own interests, you become a mentor and leader rather than a slavedriver, because you're showing them how to more effectively pursue what they already know they want.

For example, I knew someone who was trying to get his team to come in at 9:00 because the rest of the business was in at that time. If he put it in the right terms, e.g. "coming in at 9:00 will improve the image of our group and serve us better in getting promotions", he could have gotten what he wanted. But he pulled the "because I say so" / "fuck you, I'm your boss" card and flopped. If you manage from authority rather than leadership, the good ones leave you and the bad ones bring you down through incompetence if not outright sabotage.