Собеседование аналитика в RuTube

О компании

RuTube — crown jewel Russian video platform. Growing после YouTube restrictions. Part of Газпром-Медиа.

Analytics команды: content, recommendations, advertising, engagement.

Процесс

3-4 rounds:

1. HR
2. SQL / Python
3. Product case
4. Final

Стек

• ClickHouse (video events scale)
• Python
• Internal BI tools
• Spark / Hadoop для big data

SQL

Middle. Video / streaming specifics:

• Views / sessions
• Watch time
• Completion rates
• Engagement patterns

Example

«Top-10 videos by completion rate, min 10k views».

SELECT
    video_id,
    COUNT(*) AS views,
    AVG(watch_time_sec * 1.0 / video_duration_sec) AS completion_rate
FROM video_views
WHERE viewed_at >= NOW() - INTERVAL 7 DAY
    AND video_duration_sec > 0
GROUP BY video_id
HAVING COUNT(*) >= 10000
ORDER BY completion_rate DESC
LIMIT 10;

Product questions

Video platform metrics

• Watch time: главная engagement metric
• Completion rate: % видео досмотрены
• CTR на thumbnails
• Sessions per user
• Returning users

Two-sided

Creators + viewers:

• Creator retention
• Monetization (ads, premium)
• Content gaps (what topics демand?)

Recommendation

Core product. A/B recommendation algorithms.

• CTR на рекомендации
• Watch time post-recommendation
• Diversity (не только same category)

Case

«Views упали на кategories Х»:

• External events?
• Creator departed?
• Algorithm change?
• UX change?

«Как measure recommendation quality?»

• Multiple metrics: CTR, watch time, thumbs up, return rate

Growth context

RuTube активно растёт как альтернатива YouTube. Unique opportunities:

• Creator acquisition
• Content library expansion
• Monetization model evolution
• Platform differentiation

Зарплаты

• Junior: 100-150k ₽
• Middle: 150-230k ₽
• Senior: 230-350k ₽

Mid-market tech level.

Подготовка

SQL

ClickHouse strong. Scale-handling.

Domain

Video / streaming platform basics.

Product

UGC platforms (YouTube, TikTok study).

Statistics

Rank-based metrics, recommendation evaluation.

Связанные темы

• Собеседование в Kinopoisk
• Собеседование в ivi
• Recommendation systems
• ClickHouse vs PostgreSQL

FAQ

YouTube опыт?

Study as background. Not required formally.

Remote?

Partial.

ML background?

Recommendation systems — advantage.

---

Готовьтесь — откройте тренажёр с 1500+ вопросами для собесов.