Why 99% Fast Systems Still Feel Slow | Performance, Latency, Percentiles Explained(Part1)

Why do systems that are 99% fast still feel slow to users?
In this video, we go step by step into performance analysis, latency, queues, threads, and percentiles to explain a problem that most engineers misunderstand.
This is not theory-heavy fluff — this is real-world system behavior, explained with practical intuition and production examples (including Amazon-style systems).
Whether you are a backend engineer, system designer, or preparing for interviews, this video will completely change how you think about performance.

LINK TO REVISE THE CONCEPTS CHAPTERWISE - https://github.com/Sweta-ClarityNConf...


Chapters
00:00:00 Understanding Performance Analysis
00:01:29 Throughput – What It Really Means
00:01:59 Response Time Explained
00:04:07 Queueing Delay – The Real Killer
00:05:52 Concurrency vs Parallelism
00:07:24 Time Slicing Explained
00:07:37 Context Switching Cost
00:11:29 Latency vs Response Time
00:13:23 Why Average Response Time Misleads
00:14:00 Percentiles – The Right Metric
00:14:07 p50 (Median) Explained
00:16:48 p99 – Why 1% Matters
00:21:34 p99.99 – Extreme Tail Latency
00:22:54 How Percentiles Are Actually Calculated
00:25:11 Time Windows in Monitoring Tools
00:25:53 1-Minute Analysis – Spikes vs Reality
00:27:56 1-Hour Time Window – Stability vs Noise
00:29:22 Amazon Case Study – Should We Optimize p99.99?
00:30:35 Common Causes of Slow Requests
00:33:18 Question – Should We Improve 99.99%?