Source: [Stop Nodding in Meetings: How to Disagree With Your Boss and Keep Your Job](https://shrinath.me/blog/how-to-disagree-with-leadership/)

# Stop Nodding in Meetings: How to Disagree With Your Boss and Keep Your Job

Saying 'sounds good' to a plan you think will fail is not being a team player. It is withholding the thing you are paid for. How to bring your real opinion to founders and managers, clearly and early, without becoming the difficult one.

By [Shrinath Prabhu](https://shrinath.me/), Senior Staff Engineer at Avail  November 5, 2026  6 min read

You know the meeting. The founder shares the plan. The deadline is impossible, or the approach has a hole in it that you can see from across the room. Someone asks “any concerns?”

Silence. A few nods. One brave “sounds good”.

Then the call ends, and the real meeting starts in a private chat: “This is never going to work.” Everyone agrees. Nobody said it to the one person who could change it.

Six weeks later it does not work, and the founder asks the most reasonable question in the world: “Why did nobody tell me?”

This post is about getting out of that silence.

## The bot problem

Ask a chatbot whether your idea is good and it will tell you it is a great idea. It agrees because agreeing is easy. That is what makes a bot a bot: it gives you back what you put in.

Plenty of capable engineers behave the same way with people above them. Take the ticket, say yes, deliver exactly what was asked, mention nothing. It feels safe and polite.

Here is the uncomfortable part. If all you return is what was requested, you are doing a job that software now does for a few dollars a month. A senior person is paid for their judgment. Keeping it to yourself is not humility. It is not doing the job.

## Why smart people stay quiet

The reasons are real, so let me name them fairly.

-   **“It is not my place.”** It is. You are closest to the work and you can see things they cannot.
-   **“They must have thought of this already.”** Often they have not. Leaders work from summaries. The detail that breaks the plan is usually in your head, not theirs.
-   **“I do not want to seem negative.”** There is a difference between negative and useful, and it is all in the delivery. More on that below.
-   **“Last time I spoke up, it went badly.”** Sometimes the environment truly punishes honesty. That is worth knowing, and I will come back to it.
-   **Culture and upbringing.** Many of us were raised to respect seniority by not contradicting it. I understand that. But at work, the respectful thing is to help the person succeed.

## What leaders actually want

Having led a team for a while, I can tell you what it is like on the other side. You are making decisions with partial information and you know it. The person who says “I think there is a problem here, and this is why” is giving you a gift. The person who stays silent and is later proven right is no help at all.

Nobody has ever been promoted for saying “I knew it would fail”.

## How to do it well

**Start from the shared goal.** “We both want this launched by March” puts you on the same side of the table. Disagreement about the route is easy once you agree on the destination.

**Be specific.** “I am worried about this” is a mood. “The payment provider’s review takes ten working days, which puts us past the launch date” is information. Bring the second kind.

**Bring an option, not just an objection.** “This will not work” hands the problem back. “This will not work as planned, but here are two ways it could” moves things forward. You do not need a perfect answer. A starting point is enough.

**Talk in their currency.** A founder may not care that the code is messy. They care about the date, the money, the customer and the risk. Translate. “This is technical debt” lands badly. “Every feature after this will take about twice as long” lands.

**Say it early.** A concern on day two is input. The same concern in week six is an excuse.

**Pick the room.** Praise in public and disagree in private is a good default with someone senior, especially for a big objection. If it has to be in the meeting, ask it as a question first: “How are we handling the review period?”

**Ask before you assert.** Sometimes they know something you do not. “What is driving the March date?” may reveal a constraint that changes your mind. If it does, good. You were not trying to win.

**Write it down.** For anything significant, a short written note: the concern, the evidence, the options, your recommendation. Writing removes the emotion and gives them time to think. It also means your position is on record, calmly.

## Words you can actually use

-   “Before we commit, can I flag a risk?”
-   “I see it differently. Can I walk you through why?”
-   “I can do that. Here is what it will cost us.”
-   “If the date is fixed, here is what I would cut.”
-   “I might be missing context. What am I not seeing?”
-   “I disagree, but it is your call, and I will make it work.”

Notice that none of these is rude, and none of them is “sounds good”.

## Then commit

This is the half people forget. Once you have been heard and the decision goes the other way, you back it fully. No sulking, no slow-walking, no “I told you so” saved for later.

If you argue well and then commit well, something useful happens. People learn that your yes means yes and your concerns are worth hearing. That reputation is worth more than winning any single argument.

## How not to do it

-   **Every hill.** If you object to everything, people stop listening. Save it for what matters.
-   **The ambush.** Raising a major objection for the first time in front of the whole company.
-   **The hallway campaign.** Agreeing in the room and lobbying against it afterward. This destroys trust faster than anything.
-   **Making it personal.** Criticize the plan, never the person.
-   **Complaint without a path.** A problem with no suggestion attached.
-   **Reopening settled decisions** without new information.

## When the room punishes honesty

Sometimes you do all of this well and still get shut down every time. That tells you something about the place, not about you. A company that cannot hear bad news from the people closest to the work will keep being surprised by it.

You can try a different format, such as writing instead of speaking, or one to one instead of in a group. You can find an ally who has the leader’s ear. And if nothing changes, it is reasonable to decide that this is not where you will do your best work.

## The real point

AI can now produce the output. What it cannot do is look at a plan and say “I think this is a mistake, and here is what I would do”. That takes context, care and a bit of nerve, and it is a large part of [what remains of the job](https://shrinath.me/blog/ai-will-replace-coders-not-software-engineers/). It is also most of [what a staff engineer does all day](https://shrinath.me/blog/what-a-staff-engineer-actually-does/).

Next time someone asks “any concerns?” and you have one, say it. Start with “can I flag a risk?” It gets easier after the first time.

## Questions people ask

### How do you disagree with your manager respectfully?

Start from the goal you share, state your concern as a specific risk with evidence, offer an alternative, and make clear you will support the final decision. Raise it early and directly with the person, not afterward with colleagues.

### What does disagree and commit mean?

You state your disagreement clearly while the decision is open. Once it is made, you put full effort into making it work, even if it was not your choice. It only works if the disagreement was actually heard first.

### Is it risky to push back on senior leadership?

Less risky than staying quiet. Leaders generally remember who warned them about a problem and who let them walk into it. The risk comes from how you do it: complaining, being vague or reopening settled decisions damages trust, while clear, early, well-reasoned objections build it.

---

[Agent docs](https://shrinath.me/agents/index.md) · [llms.txt](https://shrinath.me/llms.txt) · [All pages](https://shrinath.me/sitemap.xml)
