Practice · carbon.talent

Analytics Engineer Interview Prep

Last updated 18 Oct 2026

Analytics engineer interviews combine SQL, data modelling, dbt practice and judgement about working with analysts. Here are original practice questions with notes on strong answers. These are original practice questions written for this page. They are not taken from, and do not preview, any assessment an employer or carbon.talent uses.

Create your free profile Browse data jobs

Modelling

  1. Design tables for an e-commerce business so analysts can report revenue by product, customer and day. Which are facts, which are dimensions, and what is the grain of each fact?
  2. A customer's address changes. How would you model it so historical orders keep the address at the time of purchase?
  3. Two teams define 'active customer' differently. How would you handle that in the modelling layer?

Strong answers state the grain explicitly, discuss slowly changing dimensions and explain how definitions are agreed and documented.

dbt practice

  1. When would you make a model incremental, and how would you handle records that arrive late?
  2. What tests would you add to an orders model, and which failures should block a deployment?
  3. How would you run only the models affected by a change in CI?

Interviewers look for real project experience: sources and freshness, generic and singular tests, state-based selection and sensible materialisation choices.

SQL

  1. Write a query that flags duplicate events with the same user, event type and timestamp, keeping the earliest ingested row.
  2. Given daily subscription snapshots, calculate monthly recurring revenue movements: new, expansion, contraction and churn.

Explain how you would validate results against finance figures.

Scenarios

Expect questions such as: a dashboard owner says yesterday's figures are wrong; what do you do? A strong answer covers checking freshness and test results, tracing lineage, communicating status, fixing and preventing recurrence.

Show your work

A public dbt project with clear layers, tests, documentation and CI is the most persuasive preparation. Be ready to explain why you structured it that way.

Common questions

What do analytics engineer interviews test?
Dimensional modelling, dbt practice, SQL and judgement in stakeholder scenarios.
Do I need a public dbt project?
It is not required but is persuasive, especially with tests, documentation and CI.
Are these carbon.talent's assessment questions?
No. They are original practice questions and do not preview any employer or carbon.talent assessment.
How do I prepare for modelling questions?
Practise stating the grain, choosing facts and dimensions and handling changing attributes.