Suppose you find yourself in the following situation:

You have a good relationship with your direct boss, and you are a subject expert in the field. The boss is telling you to do something (a technical decision), but you feel very strongly that it is a bad idea and will hurt both of you in the long term. You suspect he has been out of touch with code for a while, doing managerial things, and is not making the best decision. You suspect he is shooting himself in the foot but you're just unable to convince him of that (discussions have deadlocked). If the whole thing doesn't work later there will be repercussions for the entire team.

So the choice is to either always do what you're told, or quietly do what you think is right.

On one hand, doing the former will make sure that no blame falls on you if things go south, and if (when) it does, you will have the "I told you so" rights. On the other hand, I really don't want this guy to get screwed, and I don't want the team to spend weekends fixing production problems. This guy is a good guy, but he's not a developer anymore and he is having a hard time letting go. What's the right way to handle this?


Thanks everyone for replies, they seem more or less unanimous. I will follow your advice.

For the record, my difficulty was in the fact that we are on very friendly terms (outside of work too) with my boss. It's more than just a pure work relationship, and that is convoluding things. Shouldn't matter though - thanks everyone.

  • 2
    This is a great question. I actually decided not to accept a full time position at a company I mostly enjoyed working at because of this problem. Commented Oct 4, 2012 at 19:52
  • Wait... friends outside of work may be the problem. Your boss may feel as though if he lets up on you that he will be giving you 'favor' in the company. You may have to talk this out in depth with him to get it resolved. Commented Oct 5, 2012 at 17:36
  • 2
    If you don't do what you is told, AND it breaks, you might find yourself in a very unpleasant position. Commented May 30, 2013 at 14:15

8 Answers 8


Never quietly do what you think is right, against specific instructions from your boss. Never.

Best case is that you're right and no one ever finds out. Worst case is that you were wrong all along and then you're in serious trouble. Forget "told you so" rights, we're talking about disciplinary rights.

But I don't necessarily advocate "always do what you're told" either.

Explain to your boss that you feel very strongly about this and would like to discuss it some more. Make it very clear that if you're sure that he's understood your argument completely, you'll go along with his instructions.

Gather your data well before that meeting. Prototypes are good but don't spend too much work time on it. Or get permission first. Blogs can be useful, but should at least explain that the alternative was tried and failed, including why it failed. Ask the question on Programmers.SE and see if you get a HEAVY bias for one case over the other. Again, look for people with battle-scars.

Beyond that, argue with logic and business sense (particularly financial), not emotion.

Then, and only then, if he still tells you to just do it ... just do it. Your job is to advise, not rebel. Remember that if it all goes wrong, his neck is on the line, no matter who made the decision. So ultimately it is his call.

  • 2
    Although I hate to say this I would also advocate after the discussion sending a followup email outlying your thoughts and his decision to help safeguard you from any potential future fallout etc A polite professional email of course :)
    – dreza
    Commented Oct 4, 2012 at 19:52
  • @dreza: I would recommend that only if you were planning to invoke the "told you so" rights. I am far too old and grizzled for those schenanigans nowadays.
    – pdr
    Commented Oct 4, 2012 at 19:59
  • You have to be careful as to not step on the boss' toes however. If he's already spoken about the issue more than once, and then comes back more strongly, he may receive an adverse reaction. The boss may feel as though his authority isn't being followed and may even take offense. Commented Oct 8, 2012 at 13:31
  • @Mechaflash: Given the claim that the manager is a "good guy", and that their relationship is good, I wouldn't worry too much about that. A good manager will never close down a discussion, no matter how much extra evidence you produce. A bad manager might, but that manager will also be offended by your spending time prototyping to prove him wrong.
    – pdr
    Commented Oct 8, 2012 at 13:42
  • "Never quietly do what you think is right" - well, folklore.org/…
    – magma
    Commented Oct 23, 2019 at 13:37

You've told this person that he/she is wrong and you have not been persuasive enough to change their mind. If you believe this strongly and are willing to take on a bit of risk to help all parties involved (you, your boss, and your company), you need to show your boss that you are correct. Here are a couple ideas:

  • Build a small prototype - In the technology field, showing someone could require some time, such as building a small prototype or system that proves your point. Perhaps there are some quantifiable elements to what you could show, such as speed or scalability, in a small model of the system.
  • Call in a reinforcement - Are there other SME's that you can recruit to help you make your case (other team members, 3rd parties).
  • Make it clear that you are willing to take the risk - If you feel this strongly it seems you are willing to take on some risk of failure (if the boss agrees with your idea, at that point the risk transfers at least somewhat to you). Make it clear to your boss that you are willing to put yourself on the line with this decision and that you are potentially taking considerable risk by recommending this change. The safe bet would be to do what you are told, so let the boss know that you feel strongly enough about this endeavor that you are willing to put yourself at some risk by offering an alternate solution.
  • Research/case studies - Are there any documents or case studies available? Agree to post the question to Stack Overflow perhaps and see what kinds of answers you get?
  • Put it in the boss's best interest - When trying to sell an idea like this, you need to put the result in the boss's best interest. Get the boss to imagine the credit received for the success of the project, and the alternative results if things are not successful.
  • Time permitting, offer to do it both ways - Depending on how time-sensitive this is, would it be possible to offer a solution where perhaps you spend a specific amount of time doing it your way and another equal amount of time doing it the boss's way. Then see which partial effort makes the most sense (you want to fail fast in this case).

Good luck.


It is difficult, if not impossible, to give a complete answer to this question.

First of all, no one has complete knowledge of a situation or an issue (including domain experts). This means that quite frequently, both parties are only going to be partially correct in their solution.

Your question has the implicit accusation that your manager is out of touch. That could be, or you could be more junior than you think. Or both could be true. How can you be able to tell the difference?

Before you go after your boss, make sure you know pretty damn well that you are 100% right and he's 100% wrong. Intuition is not enough here. He will likely want to know a precise measurement of how much either choice will hurt your company.

If you want to display seniority, you provide a cost/benefit analysis. Pushing your own (possibly pet) idea will certainly not help you get across as constructive. You need to prove your point.

Secondly, please remember that you don't ever get a "I told you so" moment. If things go wrong, it will only mean that one of the two ideas didn't work. It won't prove the other idea, at all.

Finally, please don't do stuff against your boss' better judgement. That will hurt you, and possibly your company. If you are not prepared to follow the instructions of your boss, go work somewhere else! :-)


You suspect he is shooting himself in the foot but you're just unable to convince him of that (discussions have deadlocked).

Then it's my problem. As much as I hate to say it, if it is really that bad of an idea you should be able to prove it. Prototypes help. Links to (new) best practices help. Hypothetical discussions can help. Mostly, being a better advocate for your idea helps.

In the end, they're the boss. A good boss will listen and be open to how things should be done. No amount of technical debate will fix a bad boss from causing people grief in the long time. But even a bad boss is your boss. Do what they say, and work to lessen the impact of the bad idea.


I'm in the same exact situation with a boss who last coded in BASIC.

It's hard to get through to someone who is trapped in their old ways. The biggest thing that distracts them from understanding your point of view is the understanding of the method that you're trying to implement.

9/10 cases where I've brought up alterations or changes to code, or building applications under new platforms, my boss was unwilling to be flexible due to the lack of understanding of the new methodology/technique/language. I had to painstakingly walk through every minor detail of code, what purpose it posed, why that purpose was important/profitable, and what other things we can circumvent by using that code (i.e. we remove so-and-so errors, this function becomes deprecated so it's one less process to worry about, etc.).

Use diagrams and images, if possible, to portray the benefit of its use.

The utmost BEST thing to do (now this only works if time permits) is create a test environment and run a live exercise with the changes you want to make to show, in real time, what it does and why it's better.

At the same time ask yourself if what you're doing isn't shooting yourself in the foot. Sometimes we tend to believe that since we have newer knowledge that we are the better animal. However, there is still a lot you can learn from learning how they did it in the golden years ;)


The time to dispute a decsion is before it is made. In his case, you need to document the issue and prove to him why your solution is better. Do a formal cost benefit analysis. Do a formal risk analysis. Talk in business terms not just technical terms. Make sure all these analyses are in a document that has been emailed to him. If you have sent himt he information formally, he will know that if he ignores you and the risk doesn't work, you can prove it is his fault and that alone will help him to strongly consider what you have said.

If he still decides to go the other way, then you will have to do that. The decision is rightly his not yours. If you can't implement his decision, you need to quit. If your boss finds out you have gone against what he told you to do, you will be fired. And rightly so. He is making decisions considering factors you may not be aware of. Doing what you want instead of what you were directed to do is narcissistic and unprofessional. It is something no organization will tolerate for long.


You have to be able to spell out the consequences very specifically in a single sentence.

"I've seen/been-trained/have benefited from best practices that say otherwise in the past and I'm worried that if we continue down this road it's going to result in a maintainability nightmare because specific thing X will happen to specific aspect of code Y which could result in us working weekends to do specific horrible kludgey thing Z."

If that doesn't provoke conversation then I'd personally be a bit offended and start looking for new work if the situation is actually that severe. Just make sure you know what you're talking about and you're not blindly following some best practice he might be convinced is snake oil.


You should listen to your boss, because you're not paid to tell the boss what to do. Just make sure that every one knows what your feelings/advice is before starting to work on it, so your colleges wont blame you for anything if something goes wrong.

Can you imagine if the people responsible for building a nuclear submarine ignored their bosses advice and built a space shuttle, because they thought that it would travel much faster? If the boss specifically told the workers that the submarine had to be fast and hard to detect, they probably wouldn't want one desired property to be compromised over the other because both are important in this case.

  • 3
    "your not paid to tell the boss what to do" -- this seems to completely ignore that question is asked from a perspective of a subject expert in the field -- their job is often to give advice on their domain
    – gnat
    Commented Mar 7, 2014 at 6:11

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .