Most tools ask you to sign up before you can see what you get. Here is the whole thing instead. This is a composite written for this page — not a real customer’s session.
The role in this example
Data Analyst, mid-level, at a company whose posting mentions “partnering with commercial teams” three times. That repetition is why two of the six questions are about persuading people rather than about SQL.
Question 1 — Why this job?
“I have spent two years building reporting that nobody read, and this posting talks about sitting with the commercial team.”
Specific, slightly self-critical, and tied to a phrase in their advert. This is a strong opening and you should keep it almost word for word.
“…and it seems like a good company.”
This sentence does nothing. It is the kind of line people add when they are not sure they have finished. Stop one sentence earlier.
Cut the last sentence. Your answer was already done.
Question 2 — Tell me about a time your analysis changed a decision.
“I pulled the last six months of ticket data before I said anything in the meeting.”
You led with evidence rather than with an opinion. That is exactly the instinct this job is advertising for.
“…and eventually they came around.”
This is where the answer collapses. “Came around” is not a result, and you never said what convinced them. The interviewer wants the argument, because the argument is the transferable skill.
Name the specific thing that changed their mind, then say what happened next.
“I pulled six months of ticket data, showed that [ADD YOUR REAL NUMBER] of escalations came from one checkout step, and they approved the rebuild that quarter.”
You did not give us a number, so we left a bracket. We will not put a figure in your mouth that you might have to defend in the room.
Question 3 — Follow-up: what did they actually object to?
You got a follow-up here because your first answer described a situation and a task but skipped the action. This is the single most common gap we see, and it is the reason interviewers ask “and what did you do?”
“They thought the data was from a bad month.”
Good — this is the actual objection, and naming it makes the story real. This belongs in the main answer, not in a follow-up.
Question 4 — A judgement question from the posting
“Your numbers say one thing and a senior stakeholder’s intuition says another, two days before a launch. What do you do?”
“I would present the data and let them decide.”
This is a non-answer, and it is the second time you have avoided saying what you would actually do. A judgement question is asking you to commit to something.
Pick a position and defend it. “Two days out I would not block the launch, but I would write down the specific metric I expect to move and check it on day three” is an answer. “I would present the data” is not.
Question 5 — Working with people
“I started sending the weekly numbers with two sentences at the top saying what I thought they meant.”
Concrete, small, and clearly your own initiative. Short answers like this are often the strongest ones in an interview.
Question 6 — Anything you want to add?
“No, I think that’s everything.”
You had two strong stories you did not use. This question is free ground and you gave it away.
Your plan before the real interview
- Put a number on the ticket-data story. Go and find the real figure. It is the best story you have and it currently has a hole where the result should be.
- Decide your answer to the judgement question. You avoided committing twice. Pick a position now, out loud, before somebody asks you in the room.
- Prepare one thing for “anything to add?” One sentence. You will be asked it and you will otherwise say no again.
What the pattern would say after three sessions
Across eighteen answers you have given a measurable result four times. When asked what you would do, you have described a process instead of making a decision in five out of six chances. Your openings are consistently strong.
That is the thing a single session cannot tell you, and it is why the product remembers.