Most beginners should start with manual testing and move to automation once the basics feel easy. Manual testing teaches you what to test: writing test cases, reporting a bug clearly, thinking like the person using the app. Automation runs those same checks with code, so it only pays off once you know which checks matter. If you already code comfortably you can get to automation sooner, though you still need the manual skills underneath. The order we suggest is a few weeks of manual work, then one programming language, then one automation tool.
What manual testing and automation testing actually mean
Manual testing is a person using the software by hand and checking that it behaves as it should. In automation testing, a person writes a small program, called a test script, that does those checks and reports what passed and what failed.
Take a simple login form: an email box, a password box and a Login button.
A manual tester works from a written list of test cases. A test case is one check with clear steps and an expected result, for example: enter a correct email and a wrong password, press Login, and expect the message "Incorrect password" with no login. The tester also tries things the developer may not have thought about. What happens with an empty email, or spaces at the end, or a very long password? When something goes wrong, they write a bug report so the developer can see the problem and fix it.
An automation tester writes code that opens the browser, types into the same boxes, presses the same button and checks for the same message. Once it is written, the script can run on every new version of the app, on several browsers, without anyone clicking, which is the reason teams bother to automate at all.
Why most testers learn manual testing first
Automation doesn't decide what to test. A person does. If you can't write a good test case by hand, your scripts will be fast and still check the wrong things. Manual testing builds three habits that every QA (quality assurance) job expects.
- Writing test cases. You learn to break a feature into small checks, including the unhappy paths: wrong input, slow internet, someone double clicking Pay.
- Writing bug reports. A good one has a short title, exact steps, what you expected, what actually happened and a screenshot. Developers fix clear reports faster, and interviewers often ask you to write one on the spot, so practise.
- Thinking like a user. After a while you notice confusing buttons, missing error messages and screens that break on a small phone. A script won't find those on its own unless someone told it to look.
Along the way you pick up the tools teams use every day, like Jira for tracking bugs, and terms like regression testing (re-checking old features after a change). All of it comes with you when you start automating.
When to move on to automation
You're ready when you can take a feature you have never seen, write a sensible set of test cases for it and file clear bug reports without much help. For most beginners that takes a few weeks of steady practice on real websites and apps.
Don't wait until you feel perfect at manual testing. Nobody does. Plenty of people learn the two side by side after the first month, with manual testing on weekdays at work or in class and a little code each evening.
What you need to learn for automation
Automation needs one programming language and one testing tool. You do not need a computer science degree. You need enough code to write steps, store values, use loops and conditions, and read an error message without panicking.
On the tool side, two names come up in most web testing job ads: Selenium and Playwright.
- Selenium WebDriver is the older one and very widely used. Its official documentation shows code in Java, Python, C#, Ruby, JavaScript and Kotlin, and it drives the major browsers, including Chrome, Firefox, Edge and Safari.
- Playwright is newer. According to its official docs it supports Chromium, WebKit and Firefox on Windows, Linux and macOS, and you can use it from Node.js (JavaScript or TypeScript), Python, Java and .NET.
Pick one language and one tool and stay with them for a while. New learners often pair JavaScript or TypeScript with Playwright. Java or Python with Selenium is common in older projects and shows up in many interviews. Once you know one tool well the second comes much easier, since the idea is the same in both: find an element on the page, do something to it, then check the result.
Two more skills help a lot. One is API testing, which means checking the data the app sends and receives behind the screen, often with a tool called Postman. The other is Git, the tool developers use to save and share code. Most automation jobs expect at least the basics of both.
How long each takes to become job-ready
These ranges are our best guesses for someone practising most days. They aren't promises, and you may well go faster or slower.
| Stage | What you can do at the end | Rough time |
|---|---|---|
| Manual testing basics | Write test cases, file clear bug reports, use Jira, test an API in Postman | 1 to 3 months |
| Programming for testers | Write simple scripts with variables, functions, loops and conditions | 1 to 2 months on top |
| Automation with one tool | Build a small, tidy test suite for a real website that runs on its own | 2 to 4 months on top |
If you already code, the second row may shrink to a few weeks. Daily practice moves these numbers more than any one course or video does. Our post on building real projects without YouTube tutorials covers how to practise so it actually sticks.
How teams in Nepal usually mix the two
Many software companies in Kathmandu build products for clients abroad, and most of their QA teams do both kinds of testing. New features usually get tested by hand first, because they change often and a person can judge whether they make sense. Once a feature settles down, the checks that have to run again and again, such as login, payment and sign-up, get automated, and those scripts run whenever developers push new code.
Smaller companies and startups often hire one tester who does everything, so manual skills count for even more there. Larger teams sometimes split the work into manual QA roles and automation roles. Read a few local job ads and you will see that many automation posts still ask for strong manual testing and bug reporting.
We wrote separately about how QA automation training turns testers into engineers, if you want to see what changes once a tester moves into automation work.
A simple three-stage learning path
- Stage one, manual foundations. Learn how software is built and tested, write test cases for apps you already use, report the bugs you find, and get comfortable with Jira and Postman. Try to test one real public website properly. Keep those test cases and bug reports, because they are your first portfolio piece.
- Stage two, one language: JavaScript, TypeScript, Python or Java, up to functions, loops, arrays and objects. Write tiny scripts every day. When something breaks, read the error message carefully instead of copying a fix blindly.
- Stage three, one tool and one real project. Pick Playwright or Selenium, automate the test cases you wrote in stage one, keep the code organised and put it on GitHub. In an interview, a small working project you can explain does more for you than a long list of tools.
How our QA course follows this path
Our software quality assurance course runs for 8 weeks, Monday to Friday, two hours a day, in class at Old Baneshwor or online, and it goes in the same order as this post. The first four weeks cover testing fundamentals, test design, bug reporting with Jira, and API testing with Postman plus SQL. Automation starts in week five with TypeScript from zero and Playwright. After that come the Page Object Model (a tidy way to organise test code), CI/CD with GitHub Actions and performance testing, and Selenium WebDriver is taught as a second framework. You need no prior programming experience; a short assessment before day one checks where you are starting from.
Eight weeks is the short end of the ranges above, so we expect about two more hours of self-study a day. Each batch has a maximum of 10 students in the room. The last week is a capstone demo and portfolio, and you finish with two automation projects on GitHub, one in Playwright and one in Selenium. Saarathi doesn't guarantee jobs or placements. The full week-by-week plan is on the course page, and you can compare it with our other programs on the courses page.
