How To Explain Your Thoughts

Introduction

Tech jobs don’t just require hard skills - you need to be able to communicate ideas, weigh alternatives, and justify decisions too. That’s why technical interviews focus not only on the correctness of your solution, but also on how you got there.

Explaining your thought process helps the interviewer follow your reasoning, give early feedback, and assess how you break problems down. Even if you’re unsure, thinking out loud helps demonstrate logical thinking.

Key Vocabulary

Before we get started on what to explain, let's look at some key vocabulary to understand for technical interviews.

What to Explain

In a technical interview, it’s important to make your thinking visible.
This may feel unnatural, but you should explain almost everything you are thinking.

For example:

• What is the first solution that comes to mind?

• What other approaches are possible?

• Why are you choosing this approach instead of another?

• What trade-offs are you considering?

• If you are stuck, which part is causing difficulty?

Phrases

Instead of only using simple connectors to explain your thoughts and actions (“first I will”, “next I will”), try:

• I’m thinking of starting with a simple brute-force approach.

• I’ll walk you through my approach.

• My first thought is to loop through the entire list.

• Another option could be using a hash map for constant lookups.

• I’m debating between a recursive solution and an iterative one.

• Instead of looping twice, I can just loop once.

Structure

If you find yourself going silent while live coding or thinking over a solution, try this structure:

1. Describe your initial idea

2. Mention trade-offs or alternative approaches

3. Justify your choice

4. State what you’ll do next

See this structure in action in the model answers below.

Model Answer: Data

Interviewer: We have a dataset of customer purchases. How would you identify the most popular product by region

Candidate: [Candidate asks clarifying questions…]

Okay now I understand all the details, I’ll walk you through what I’m thinking.

First, I’d want to group the data by region and product to count the number of purchases.
I’m debating between using SQL or Python for this.

Since the data is clean and stored in a database, SQL will be more efficient.
I can run a GROUP BY and COUNT() directly on the server.
Python with pandas gives more flexibility but could be slower for large volumes.

Given that the data is already in good shape, I’ll go ahead with SQL for performance.

My approach will be: group by region and product, count purchases, sort, then use a window function to select the top product per region. How does that sound?

Model Answer: Product

Interviewer: We’ve seen a decline in conversion from free to paid users. How would you investigate this?

Candidate: [Candidate asks clarifying questions…]

Okay, I’ll walk you through what I’m thinking.

First, I want to map the current user journey, from sign-up to conversion, and identify the key touchpoints.

We could start with analytics tools to analyse drop-offs, or run user interviews to uncover friction points.
Analytics are fast for spotting patterns; interviews provide deeper context.

Since we need to act quickly and already have tracking set up, I’d start with data.
I’d check cohort behavior and compare conversion rates over time.

Then I’d pull a conversion funnel, compare it to a previous healthy period, find the biggest drop-offs, and follow up with interviews for qualitative insight.