Why 900+ Interviews Taught Me to Ignore the Resume First

Interview Insights · Hiring Approach

Why 900+ Interviews Taught Me to Ignore the Resume First

SECTION 01What goes wrong when the resume comes first

A resume is a marketing document. That is not a criticism; it is supposed to present you well. But when an interviewer reads it first, it sets an expectation, and the rest of the interview gets judged against that expectation. A candidate with a famous college name or a long skills list starts with credit they have not earned in the room. A candidate with a plain layout starts behind.

The second problem is rehearsal. Every line on your resume is something you have practised explaining. When I ask about it, I hear a prepared answer. That answer tells me you can prepare, which is useful, but it does not tell me how you handle something new. Data work is mostly new problems.

Interviewer: I see you used Power BI at your internship.
Tell me about that.

Candidate: I built a sales dashboard with slicers for
region and month. It had KPI cards for revenue, orders
and average order value. The manager used it weekly.

(Smooth, rehearsed, accurate. But I still do not know
how this person reacts to a question they have not
prepared.)

Interviewer: Before we look at your resume, a quick one.
An online store's revenue fell this month compared with
last month. How would you find out why?

Candidate: Can I ask a few things first? Is the fall in
all categories or some? Did the number of orders fall,
or the average order value? And is this month complete,
or are we comparing a partial month with a full one?

(Now I am seeing how they think.)

The second candidate had not said anything impressive yet. They asked three questions that a working analyst would ask. That is more informative than any bullet point, and it came in the first minute.

SECTION 02The small problems I open with, and what they reveal

Opening problems are small on purpose. They should take five to fifteen minutes and should not need special tools. Most of them are business questions with data behind them, because that is the core of an analyst's job. I am not looking for a perfect answer. I am looking for the order of your thinking: do you clarify, break the problem down, choose a method and say your assumptions?

Here are the kinds of problems I use for data analyst roles, and what each one shows me. You will meet similar ones from many interviewers, even if they phrase them differently.

-- Customers with exactly one order, and that order
-- placed more than 60 days before the latest order date
WITH stats AS (
  SELECT customer_id,
         COUNT(*)        AS orders,
         MAX(order_date) AS last_order
  FROM orders
  GROUP BY customer_id
)
SELECT customer_id, last_order
FROM stats
WHERE orders = 1
  AND last_order < (SELECT MAX(order_date) FROM orders)
                   - INTERVAL '60 days';

-- Spoken assumption: 'never returned' = no second order
-- within 60 days of the data's latest date. PostgreSQL syntax.

Revenue = Orders x Average Order Value (AOV)

Step 1: Did orders fall, AOV fall, or both?
Step 2: If orders fell -> split by channel and device.
        Did traffic fall, or conversion rate?
Step 3: If AOV fell -> split by category.
        Did a high-price category shrink? Discounts?
Step 4: Check calendar effects: is the month complete?
        Did last month have a festival sale?

Spoken line: 'I would not guess a cause until I know
which of the two factors moved.'

Notice the comment in the SQL tab. The candidate states the definition of 'never returned' before writing the query. In a live interview, that one spoken assumption is often worth more than getting every bracket right.

SECTION 03Where the resume finally comes in

I do still read the resume. I simply read it after I have seen you think. At that point it becomes a tool for checking, not a source of first impressions. If you handled the revenue problem well and your resume says you built sales dashboards, the two pieces of evidence agree, and I trust both more. If you struggled with basic grouping and your resume lists advanced SQL, I ask about that gap directly.

For candidates, this means the resume still matters, but in a different way. Every line should be something you can demonstrate live, because the interviewer will compare it with what they just saw.

Interviewer: Your resume says advanced SQL. Earlier
the GROUP BY took a while. Help me understand.

Weak: 'I was nervous.' (and nothing else)

Strong: 'Fair point. I have used window functions in
my project, mostly ROW_NUMBER for deduplication, but I
rely on looking up syntax. I should have said
intermediate. Can I show you the dedup query I wrote?'

FOR EVERY LINE ON YOUR RESUME, ASK:

1. Could I demonstrate this live in five minutes?
2. Could I explain one decision I made in it?
3. Could I name one problem I hit and how I fixed it?

If the answer to 1 is no:
 -> rewrite the line at the level you can demonstrate
 -> or remove it

Example:
'Advanced SQL' -> 'SQL: joins, CTEs, ROW_NUMBER for
deduplication in a retail sales project'

The strong gap answer does something brave: it lowers the claim. That honesty usually earns more trust than defending the original line, and it moves the conversation back to something real you can show.

SECTION 04Preparing for an interviewer who starts with a problem

You cannot predict the exact opening problem, but you can prepare the habits that handle any of them. Use these three tabs over the week before your interview: a clarifying-question bank, a practice set of opening problems and a short self-review after each attempt.

CLARIFYING QUESTION BANK (pick 2-3 per problem)

Scope:     Is this for all regions / products / users?
Time:      Which period? Is it complete? Compared to what?
Definition: What exactly counts as active / returned /
            churned / a sale?
Data:      Which tables or sources do we have?
Audience:  Who will use the answer, and for what decision?
Urgency:   Do you want a quick estimate or a full answer?

OPENING PROBLEM PRACTICE SET (answer aloud, 10 min each)

1. Website sign-ups doubled last week. Good or bad?
2. Find the top three products by revenue in each city.
3. Customer complaints rose. What data would you pull?
4. Two teams report different 'active user' counts.
5. Estimate how many orders a café gets on a weekday.
6. Which chart shows monthly sales by region best?

For each: clarify, break down, approach, assumptions.

AFTER EACH PRACTICE PROBLEM

[ ] Asked at least one clarifying question
[ ] Broke the problem into parts before answering
[ ] Said my assumptions aloud
[ ] Named the data or SQL I would use
[ ] Gave a next step, not just an answer
[ ] Stayed under 15 minutes

Weakest habit today: ______________
Fix for next attempt: _____________

SECTION 05Habits that help when the resume is not your safety net

Many candidates lean on their resume as a script. These habits help you perform well even when the interview starts somewhere else.

• Expect the unexpected opening — Assume the first question may be a problem, not 'tell me about yourself', and practise switching into problem mode quickly.
• Clarify before you calculate — One good question about scope or definition is the strongest possible start to any vague data problem.
• Break metrics into factors — Revenue into orders and order value, sign-ups into traffic and conversion rate. It works for most business problems.
• Write resume lines you can demo — Every claim should be something you could show live, because it will be compared with what you just did.
• Own a gap honestly — If your live performance and resume differ, say so calmly and offer a concrete example of what you can do.

SECTION 06Interview Q&A — Resume-last interviews

SECTION 07Test yourself — Resume-last interviews quiz

SECTION 08Frequently asked questions

SECTION 09Related reads