Performance Testing Fundamentals
Part of the QA Testing Roadmap
In short
An app can pass every functional test and still fall over when the sale goes live. Performance testing asks a different question: not 'does it work?' but 'does it still work for everyone at once?'
Verifying a system performs acceptably under expected (and unexpected) load — a distinct discipline from functional testing, focused on speed, stability, and scalability rather than correctness alone.
Think of it like this
The new metro line on opening day
During trial runs with ten engineers, the metro feels perfect. On opening day, fifty thousand people show up. Gates jam, platforms overflow, trains run late. Nothing was 'broken' before. It just never met real crowds.
Watch it happen
Response time as users grow
1/5Each cell is the typical response time in milliseconds at a load level: 10, 100, 500, 1,000 and 2,000 users.
The big ideas
What you need to know
Response time
How long one request takes. Look at percentiles like p95, not just the average.
Throughput
How many requests per second the system handles. It should grow with users until it hits a limit.
Error rate
The share of requests that fail. Slow is bad. Failing is worse.
Resource usage
CPU, memory, database connections. They explain why the numbers above change.
How it works
Your first performance test
- 1
Get permission
Only load test systems you own or are explicitly allowed to test, ideally a dedicated test environment.
- 2
Pick one flow
Choose a key flow like search or checkout, not the whole app.
- 3
Set a target
Agree what good looks like, such as 95% of searches under 500 ms.
- 4
Ramp and watch
Increase virtual users step by step and record response time, throughput and errors.
Save yourself the pain
Common mistakes
Oops
Running a load test against production or a site you don't own, without permission.
Do this instead
Use a test environment and written approval. Unplanned load can take real systems down.
Oops
Reporting only the average response time.
Do this instead
Report percentiles, throughput and error rate together. Averages hide the slow requests.
Your mission
Watch performance by hand
Quick check
What does performance testing mainly try to find?
Go deeper
- 1Learn the key performance metrics: response time, throughput, and error rate.
- 2Understand why a functionally correct system can still fail under real-world load.
- 3Read about a real production incident caused by inadequate performance testing.
More topics at this stop
Load, Stress & Spike Testing
JMeter Fundamentals
k6 / Gatling (Overview)
Analyzing Performance Test Results
Identifying Bottlenecks
Performance Testing in CI/CD
Setting Performance Benchmarks
Stuck on this topic? Ask an Insider
Get 1:1 guidance from people who've walked this exact path — on the InsideEdge app.
