Siphon Effect

A siphon is a curved tube used to draw liquid out of a container. Once the tube is filled, the water keeps flowing on its own until the container is empty. This exact image lies behind the term siphon effect. It refers to the way an offering pulls users, money, or professionals away from other providers, without those providers being able to do anything about it. In business journalism, the expression tends to appear whenever a new service noticeably draws customers away from an established one. So the effect describes a redirection, not a creation of something new: the pie doesn’t get bigger, it just gets divided differently.

Why redirection can be worse than competition

Many internet services rely on people coming to visit them. A news website earns money through advertising, which only pays off if someone actually opens the page. If this stream of visitors gets redirected, the revenue source dries up even though the content itself continues to be consumed. This is exactly the painful part of the siphon effect: the service keeps being used, but the benefit ends up somewhere else.

A current example is AI chatbots that answer questions directly. In the past, people would type a question into a search engine and then click through to a website. Today, the chatbot delivers the answer immediately, often relying on that very website for its information. The site’s operator did the work but no longer sees a visitor. Publishers therefore speak of siphoned-off traffic, meaning drained-away visitor numbers.

The effect doesn’t just affect clicks. Capital and staff can be siphoned off too. When a handful of AI companies pay extremely high salaries, researchers switch over from universities and smaller firms. For the institutions losing them, this is a creeping loss that only becomes visible years later in missing results.

What keeps the pull going

For a siphon to work, it needs a gradient. In business life, this gradient is convenience. People almost always choose the path of least effort. A service that provides an answer without an extra click therefore wins out, even when a competitor’s content is actually better.

This is reinforced by the network effect: the more people use a service, the more attractive it becomes for the next ones. More users mean more data, more data means better answers, and better answers attract even more users. This feedback loop is the reason the flow rarely dries up on its own. If anything, it accelerates.

It’s important to distinguish this from plain competition. In normal competition, a provider wins because it has a better product of its own. With the siphon effect, by contrast, the winner sits in a position where it passes along other people’s work. It profits from content that others created. That’s why the term regularly ends up in courtrooms and before competition authorities.

Where the term shows up in the news

Most often, one reads about the siphon effect in disputes between media companies and AI providers. Publishers present statistics showing that traffic coming through search engines has dropped significantly since the introduction of AI-generated answers. This has led to lawsuits, licensing deals, and political debates over a right to share in the resulting revenue.

The term also plays a role in quarterly reports. Analysts regularly ask online retailers, travel portals, or comparison sites whether AI assistants are drawing customers away from them. A company that depends heavily on search hits is considered vulnerable on the stock market. Such assessments sometimes move share prices more strongly than the actual figures do.

A common misconception is that the effect only hits small providers. In fact, it can catch even very large companies off guard if user behavior changes fundamentally. Anyone who wants to understand how stable a digital business model really is should therefore always ask: who actually controls the path by which customers arrive?