Why It Matters
Every server has a ceiling. There is only so much memory, processing power, and bandwidth you can add to a single machine before you hit physical limits or prohibitive costs. Horizontal scaling removes that ceiling. When you need more capacity, you add another server. When demand drops, you remove servers and stop paying for them. This flexibility makes it the preferred approach for businesses with unpredictable or growing traffic. It also improves resilience: if one server fails, the others continue handling requests, so your site stays online. The trade-off is that your application needs to be designed to work across multiple servers, which requires more upfront planning than simply running everything on one box. Scalability covers the broader concept if you want to understand the different approaches. Browse the glossary for related definitions.
Example
A job board experiences heavy traffic on Monday mornings when employers post new roles and candidates log in to search. During the rest of the week, traffic is moderate. Rather than paying for a powerful server that sits idle most of the time, they use horizontal scaling: three servers handle the Monday morning rush, and the infrastructure automatically reduces to one server during quieter periods. This keeps the site responsive during peak hours while keeping hosting costs proportional to actual usage.
