In this article
Before you start
Why interviewers ask this
Most of the work that matters crosses team lines, and the org chart gives you no power over the people on the other side. The question tests whether you can still get a result from them, and how you behave when your opinion differs from theirs.
Google's structured-interviewing guide (updated March 2026) uses behavioral questions about past experience, such as "Tell me about a time…", which it says reveal patterns of behavior. A behavioral question is a question that asks for a real past event instead of an opinion. It describes scoring each answer against a rubric, which is a written scale of what poor, borderline, solid and outstanding answers look like. A weak answer says "I am good at working with people" and offers no event. A strong answer tells one event with an outcome. This page is built around that scale: the rubric below has the same shape.
Influence without authority from first principles
- Exchange
- An exchange is a trade in which you give something the other person values, such as data, credit or reduced scope, in return for what you need from them.
- Escalation
- Escalation is a step that takes a disagreement to a shared manager or decision-maker to settle, which a good story treats as a last resort and not a first move.
In plain words
Think of asking a neighbour to move their car. You cannot order them to. So you find out why it is there, and you offer something that suits them, such as a time when it will be easy. The analogy breaks in one place. A neighbour you may never see again owes you nothing. A colleague you will work with for years remembers how you behaved, so trust builds up or drains away with each request.
Cohen and Bradford's book (third edition, November 2017) calls this the Exchange Model: you get what you need by offering something the other person values. Cialdini (2001) argues in the Harvard Business Review that leadership is largely getting things done through others and that persuasion rests on established psychological research, not on an innate talent.
Step-by-step description
- Learn what they need.
- Offer something they value.
- Ask what would change it.
- Start by finding out what the other person needs.
- Make an exchange that serves both sides.
- Treat a no as information and escalate last.
Step 1: Find out what the other person needs
Start with questions, not a pitch. Ask what the other team is measured on, what is on its plate this quarter and what a bad outcome would look like for it. A platform team might be judged on stability, a product team on a launch date, a data team on the quality of a dataset. Each target tells you what that team will protect and what it might trade.
Input: a goal you need help with. Output: one sentence on what the other person values and what worries them. If you cannot write that sentence, you are not ready to ask. You now know what to offer.
Step 2: Make an exchange that serves both sides
State your ask in terms of their goal, make it small and specific, and offer something they value in return. "I need the schema change by Friday" is a demand. "If the schema change lands this sprint, your team's dashboard stops breaking every release, and I will write the migration tests" is an exchange. Useful things to offer include data that supports their case, credit in front of leadership, reduced scope, doing part of the work yourself, or taking an unwanted task.
Input: their goal and your ask. Output: a proposal in which both sides gain. Say what you are not asking for, because a narrow ask is easier to accept. You now know how to open the conversation.
Step 3: Handle the no
A no is information, not a defeat: ask what would have to change for it to become a yes. Often the answer is a constraint you did not know about, such as a freeze, a staffing gap or an earlier promise. Then adjust the ask, trade something else, or agree a smaller first step that proves the idea.
Escalation is the last move, and the way you do it is part of the answer. Bring the shared manager both positions in one message, framed as a decision to make, not a complaint about a person. Reported practice differs on how early to escalate; the safe reading for an interview is that the story shows you tried a trade first.
Input: a refusal. Output: a changed ask or a decision made together. This step is where answers are separated, because it shows how you behave when the first move fails.
Text description of the diagram
- A strong answer names the outcome you needed and the other person's interest.
- A Staff answer adds what you changed afterwards.
Knowing it is half the job. Saying it is the other half.
Interview Clozo is an AI interview copilot for real interview rounds: it hears the question and puts a strong answer on your screen, and you say it in your own words. See the features or how it works.
Start Free Trial →The Staff-level answer
The example below is an illustrative composite written for this page, not a real case, and its figures are made up. Replace it with your own story; the structure is what to keep.
Tell me about a time you influenced without authority.
Our service needed another team to change an application programming interface (API) before a launch, and I had no say over their roadmap. I first asked what they were measured on: stability, and a freeze two weeks before the launch. So I proposed a smaller change that kept their API stable, offered to write the compatibility tests myself, and sent their lead the launch data showing why it mattered. They agreed after one counter-proposal on timing. We launched on schedule with their change behind a flag. What I changed afterwards: I now ask for each team's constraints in the planning meeting, before I need anything.
A strong answer names the outcome you needed and the other person's interest. What makes this one Staff-level: the opening names the lack of authority, the move is anchored in the other team's goal, the ask shrinks when it meets a constraint, and the last sentence is a lasting change in how the speaker works. It runs about 40 seconds when spoken: 105 words at the 152 to nearly 170 words per minute that Yuan, Liberman and Cieri (2006) report is 37 to 41 seconds.1 A weak version of the same story is "I convinced them it was important and they did it": no goal of theirs, no trade, no result, and nothing learned.
The rubric
| Dimension | Weak answer | Strong answer | Staff answer |
|---|---|---|---|
| Situation | Vague, no stakes | One event with a clear need | Event shows why authority was missing |
| Understanding them | Assumes what they want | Asks what they need | Ties the ask to their measured goal |
| The move | Persuades or pushes | Offers a trade | Shrinks the ask to fit a constraint |
| Handling a no | Escalates or gives up | Adjusts the ask | Trades, then escalates as a joint decision |
| Result and learning | None, or only "it worked" | A concrete result | A result and a change in how you work |
Three variants of this question
How do you push development when you do not have the explicit support of the rest of the team?
Treat the missing support as the first problem. Find the one person with the most to lose from the status quo, learn their goal and make the smallest change that proves value. Evidence and a small trial usually move a team faster than an argument; see engineering management for the team-leading side.
Describe a time you influenced a peer who had a differing opinion about a shared goal.
Start from the shared goal and say the disagreement was about the route. Show that you restated the peer's view until they agreed it was accurate, then proposed a test or a decision rule that both could accept. Ending on the shared goal keeps the peer on your side.
Tell me about a time you influenced people to reprioritize their work.
Reprioritisation means someone must give something up, so say what you traded for it. Name the cost to them, the benefit that outweighed it and who agreed to the new order. Leadership interviews go further into asking senior people to move priorities.
Common mistakes
- "I convinced them because I was right." (wrong) It treats influence as winning an argument and gives the other person nothing. (right) "I learned what they were measured on and proposed a change that helped both of us."
- "I went to my manager." (wrong) Opening with escalation suggests you cannot work across a line. (right) "I tried a smaller trade first and escalated only the decision we could not settle together."
- A story with no result. (wrong) The interviewer cannot tell it worked. (right) End on the outcome and what you changed afterwards.
- "We all worked together." (wrong) It hides your own part. (right) Use "I" for your moves and name the other side's contribution.
Follow-up questions
What if the other person still says no?
Ask what would have to change for a yes, and adjust the ask or trade something else. If it still fails, bring the shared decision-maker both positions in one message and accept the outcome. A story that ends honestly with a partial result often reads better than a tidy win.
How long should the story be?
About 100 to 200 words, which is roughly 35 to 79 seconds at the same 152 to nearly 170 words per minute (Yuan et al., 2006),1 with the outcome in the first third so the interviewer knows where it is going. Save the detail for follow-ups. The tell me about yourself page has a similar timing rule.
How is this different from leadership?
Leadership questions often assume a team you lead; this one assumes none. The skill is the same: understand what people value and align work to it. See the interview loops page for where each kind of question appears.
Does this question come up outside engineering?
Yes. Product, data and consulting roles all depend on people they do not manage; the product management hub covers the product version. Change the stakeholders in the story, not the structure.
- Influence without authority is getting action by understanding what the other person values, not by rank.
- Learn their goal first, then offer a specific exchange, then treat a no as information.
- Escalate last, and bring both positions as a decision to make.
- Put the outcome and what you changed in your answer; the greatest weakness question uses the same closing move.
Sources
- Towards an integrated understanding of speaking rate in conversation (Yuan, Liberman and Cieri) — Interspeech 2006 (Linguistic Data Consortium), 17 Sep 2006
- A guide to structured interviewing for better hiring practices — Google re:Work (updated March 2026), 1 Mar 2026
- Influence Without Authority, 3rd edition (Cohen and Bradford) — Wiley-VCH (publisher listing, November 2017), 1 Nov 2017
- Harnessing the Science of Persuasion (Cialdini) — Harvard Business Review (October 2001), 1 Oct 2001
