Viber Data Analyst Interview: Questions & Prep (2026)

Viber is a global messaging and voice calling platform with a strong product and data focus. As of July 2026, knok's job radar shows Viber has 36 open roles across its tech and data teams, reflecting active hiring. Data Analyst roles at Viber sit at the crossroads of product, growth, and engineering: you are expected to make sense of messaging behaviour, user retention, and feature performance at scale.

Candidates report the process typically includes a recruiter screening call, followed by a take-home case study or an SQL and analytics assessment, and then one or more panel interviews with data leads or product managers. The exact number of rounds varies by team and level, so confirm the format with your recruiter early.

Knok's job radar salary data for Data Analysts in India:

Individual offers vary based on skills, negotiation, and the specific team. These ranges align with Glassdoor and levels.fyi figures for product analytics roles at global tech companies with Indian operations.

Viber's interview questions tend to focus on product analytics, SQL problem-solving, and how you communicate data insights. Here are questions candidates commonly report encountering:

1. A core Viber metric, daily active users, drops suddenly overnight. Walk us through how you would diagnose what happened.
2. How would you design a reporting dashboard to track message delivery rates and read rates across different geographies?
3. Tell us about a time you worked with messy or incomplete data. How did you handle it and what was the outcome?
4. What metrics would you use to measure the success of a new Viber feature, such as a group calling upgrade?
5. How do you approach setting up an A/B test on a messaging platform where users influence each other's behaviour?
6. Explain the difference between a data warehouse and a data lake. What would you recommend for a platform like Viber and why?
7. How would you build a model or ruleset to identify users at risk of churning from Viber?
8. A product manager wants proof that a new in-app sticker pack increased engagement. How do you structure the analysis?
9. Write a SQL query to rank countries by average messages sent per user over the past month. How would you handle ties?
10. Viber operates across many countries with very different usage patterns. How would you segment users for an engagement study?
11. How do you communicate a surprising or counterintuitive data finding to a non-technical stakeholder?
12. How would you track notification opt-out rates and measure their effect on user retention?

Q: A core Viber metric, daily active users, drops suddenly. Walk us through your diagnosis.

*Situation:* At my previous company, our daily active user count fell sharply on a Monday morning with no obvious cause.
*Task:* I needed to quickly identify whether the issue was a data pipeline failure, a product bug, or a real change in user behaviour.
*Action:* I first checked pipeline health and confirmed that data ingestion had completed correctly. Then I broke the metric down by platform (iOS, Android, web), by country, and by user cohort. The drop was concentrated on Android users in one specific country. I cross-referenced with the engineering release log and found a push notification bug had shipped the previous Friday.
*Result:* Within a few hours the engineering team rolled back the change and user counts recovered. I documented the diagnosis steps as a runbook so the team could move faster on future incidents.

---

Q: Tell us about a time you worked with messy or incomplete data.

*Situation:* I was analysing user session data for a product team and discovered that a notable share of sessions had null or inconsistent device identifiers.
*Task:* I needed to decide whether to drop those records, impute values, or flag them separately, without biasing the analysis.
*Action:* I first investigated why the nulls existed, which revealed a tracking gap on a specific app version. I excluded those sessions from percentage-based metrics but included them in absolute counts with a clear footnote. I also raised a ticket with engineering to fix the tracking upstream so the problem would not recur.
*Result:* The product team received an accurate picture of engagement and understood the data limitation without losing confidence in the report. The tracking bug was fixed in the next release.

---

Q: How do you communicate a counterintuitive finding to a non-technical stakeholder?

*Situation:* An analysis I ran showed that users who received fewer push notifications had higher retention, which was the opposite of what the growth team expected.
*Task:* I had to present this credibly and make it actionable, without the stakeholder dismissing it as a data error.
*Action:* I walked them through the data quality checks I had run, then used a simple chart showing retention curves side by side for high-notification and low-notification user cohorts. I framed it around 'notification fatigue,' a concept widely discussed in industry surveys and publicly reported research on consumer apps, and I avoided statistical jargon throughout.
*Result:* The growth team paused a planned notification push and ran a controlled test instead. The test confirmed the finding, notification volume was reduced, and retention metrics improved in the following quarter.

The Metric Tree for investigation questions. When asked to diagnose a drop or spike in a number, break the top-level metric into its components first. For example, daily active users equals new users plus returning users minus churned users. Then identify which branch changed, and keep decomposing until you find the root cause. This shows structured thinking rather than guessing.

STAR for behavioural questions. Situation (brief context), Task (your specific responsibility), Action (what you did, in detail), Result (measurable outcome or learning). Keep Situation and Task short and spend most of your time on Action. Viber interviewers want to see your personal contribution, not what 'the team' did.

Hypothesis-first for open-ended case questions. Before diving into analysis, state your hypothesis. For example: 'I expect the drop to be platform-specific rather than global, because a global issue would also affect our web product.' This signals product intuition and prevents you from appearing to explore randomly.

The 'so what' close. For any analytical question, end with the business implication. Do not stop at 'I found that X.' Add: 'which means the team should consider Y.' This is what separates analysts who run numbers from those who influence decisions.

Strong SQL. Expect at least one hands-on SQL question. Viber's data teams work with large event tables, so be comfortable with window functions, CTEs, and aggregations across partitioned datasets. Practice writing clean, readable queries, not just queries that technically work.

Product sense for a messaging platform. You should be able to speak naturally about what makes a messaging product healthy: message send rates, delivery rates, read rates, session length, and retention cohorts. If you have not used Viber before, spend time with it ahead of the interview and form genuine opinions about its features.

Comfort with ambiguity. Candidates report that Viber interviewers often give underspecified problems on purpose. Asking clarifying questions (for example, 'Should I focus on a specific region or user segment?') is a positive signal, not a sign of weakness.

Clear communication. Data Analysts at Viber work closely with product managers and engineers who may not be data-fluent. Interviewers pay attention to how clearly you explain your reasoning. Avoid jargon when a plain word works just as well.

Intellectual curiosity. Viber competes in a fast-moving messaging market. Showing that you have thought about the space, know what metrics matter, and have opinions about product decisions will set you apart from candidates who offer only technical skills.

Week 1: SQL and data foundations. Practise SQL on platforms like LeetCode or StrataScratch, focusing on window functions and multi-step aggregations. Review how you would structure a schema for a messaging product: users, sessions, messages, and events.

Week 2: Product metrics and case studies. Study how messaging platforms measure success. Practise breaking down metrics like retention, engagement, and churn using a metric tree. Find a publicly reported analysis from a messaging or consumer tech company and walk through it out loud to build your narrative.

Week 3: Mock interviews and communication. Do a couple of mock interviews with a friend or peer. Record yourself answering behavioural questions and check whether you lead with the key point or bury it. Work on keeping STAR answers focused and concise.

Week 4: Company-specific research. Use Viber as a regular user. Note features you find interesting or unclear and form opinions on what the data might show. Review publicly reported news about Viber's product direction in 2025-2026. Prepare thoughtful questions to ask your interviewer about the team's data stack and priorities.

After each round. Send a brief, personal thank-you note to your recruiter or interviewer promptly after the call, ideally the same day. Mention one specific topic from the conversation to make it feel genuine rather than templated. Candidates report this is noticed, though practices vary by interviewer.

Knok checks 150+ job sites nightly, applies to jobs that match your resume, and messages HR for you, so you can spend your prep time on interview readiness rather than application tracking.

Jumping to a conclusion before decomposing the metric. When given a 'why did this metric drop' question, many candidates immediately guess a cause. Interviewers want to see you break the number down systematically before forming a hypothesis.

Generic or unreadable SQL. Avoid writing a query that works but is hard to follow. Use clear aliases, consistent formatting, and a brief comment if a step is non-obvious. Viber's data teams collaborate on code and readability is treated as part of quality.

Listing metrics without connecting them to a business outcome. Answering 'how would you measure feature success' with a list of metrics, without explaining which one matters most or why, misses the point. Always tie your metrics to a decision the business needs to make.

Overcomplicating the statistics. Some candidates propose complex models when a simple cohort comparison would answer the question just as well. Start with the simplest approach and add complexity only if the interviewer explicitly pushes for it.

Not asking clarifying questions. Accepting an ambiguous brief without asking anything signals that you will build analyses the stakeholder did not actually need. Treat an underspecified question as an invitation to demonstrate product thinking.

Dropping the result in STAR answers. Many candidates give detailed context and actions but trail off before the result. A result without a specific number is still a result: 'The team adopted my recommendation and changed the notification strategy' is far better than ending at the action.