Most of a buyer's evaluation now happens on your site before they ever speak to you. They read, compare, and weigh their options, and the trail they leave behind says a fair amount about where they are. Not all of it means anything, though. It is worth being precise about which behaviors actually point to intent, because reading too much into the wrong ones is its own kind of mistake.
A buying signal is any behavior that shows a visitor evaluating rather than browsing. The well-known ones, like returning more than once or spending real time on your case studies, are only part of it. A document downloaded, or real time spent on a page most visitors skim, is a small investment of effort that a casual browser does not make. More than one person from the same company arriving in a short span is another, since it usually means a decision is being weighed internally rather than one person idly looking.
Another signal sits not in how often someone visits but in the kind of pages they move toward. Early research tends to stay on the pages that explain what you are. Someone getting more serious moves to the practical detail instead: how you actually work, what an engagement involves, your terms, and the things people only check when they are weighing a real decision, like who would be on their account or how your security holds up. That move from learning what you do to looking at how it would actually work is easy to miss, and it shows someone considering working with you rather than just reading about you. It often happens inside a single session, so the direction someone takes through a visit can tell you as much as how many times they come back.
Plenty of behavior looks like intent and is not. A long time on a page can mean deep interest or a tab left open while someone made coffee. A visit from a competitor or a job seeker can look like a prospect if you only count pageviews. Heavy traffic to a blog post often reflects a popular topic rather than anyone close to buying. The honest framing is that these are signals, not proof. They tell you what someone looked at, and they are worth acting on, but they do not tell you what a person is thinking, and treating them as certainty leads to awkward calls.
All of this works on first-party data, the behavior a visitor leaves on your own site by choosing to come and read it. That is a different thing from the software that buys third-party profiles or follows people across the web to put a name to an anonymous visitor. That second kind sits on shaky ground under GDPR, because it processes personal data the person never agreed to hand over, and the consent behind a bought profile is usually impossible to show. First-party signals stay on the footing you already have with your own visitors, the same lawful basis you rely on for your CRM and your calendar. A rep acting on it opens the conversation where the visitor's attention already was, rather than starting cold.
Leadop reads a lead's own behavior on your site, weighs the signals that point to intent, and hands the rep a short briefing before the meeting, so the patterns that took place over days are visible in the seconds before a call.