What does 'scalable' mean in software?

Three Types of Growth

More users simultaneously. Ten or a thousand people at the same time make a significant difference in the architecture.

More data. A search over ten thousand data sets is trivial, over ten million it is not.

More processes. One hundred orders per day or one hundred thousand require different processing.

These three rarely grow at the same rate – and you should know which one will increase first in your case.

What Scalability Practically Means

• Clean database structure with appropriate indexes
• Resource-intensive work in the background, not while the user waits
• Caching for frequently accessed data
• The possibility of adding a second server

The Expensive Mistake in Both Directions

Too little: After two years, the system becomes slow, and the repair costs more than the original setup.

Too much: An architecture for millions of users with only twenty actual users. More expensive in construction and operation, and no one understands it anymore.

The Right Question

Not 'is it scalable?', but 'what volume are we building for in five years?' Giving an honest answer to this is more valuable than any technical assurance.