the truth about dwell time, bounce rate, and google rankings
there's a long-standing belief that google tracks how long someone stays on your page, and that leaving quickly — bouncing — actively hurts your ranking. it's repeated often enough that most small business owners treat it as settled fact. it doesn't hold up well once you actually think through how it would need to work, and it isn't listed among the actual signals in google's page experience documentation.
the basic problem with the theory
most blog posts and informational pages have a naturally high bounce rate, and that's not a sign of failure. someone searches a specific question, finds the answer, and leaves. that's the page doing its job correctly, not a visitor rejecting it. if "leaving quickly" were a straightforward penalty, nearly every well-answered informational query on the internet would be punishing its own best-performing pages.
there's also a practical data problem: on-page behaviour like time-on-site isn't reliably available to google at scale. privacy regulations vary by country, not everyone runs analytics that reports this kind of data back, and browser and ad-blocking differences mean the data that does exist is inconsistent. it's a shaky foundation to build a global ranking signal on.
what seems to matter more: behaviour on google's own results page
there is a pattern that's much easier for google to measure reliably, because it happens entirely within google's own product: pogo-sticking. that's when someone clicks a result, doesn't find what they need, hits back, and clicks a different result instead — repeatedly, across many different searchers, for the same query.
that pattern is observable directly on google's own turf, with none of the data-availability problems that plague on-page tracking. if a result is consistently getting clicked and then abandoned in favour of a different result, that's a much cleaner signal that the page isn't actually satisfying the query — regardless of how long any individual visit lasted.
what this means practically
the actionable version of this isn't "keep people on the page longer." it's "make sure people don't need to leave and try someone else." those sound similar but lead to very different decisions:
- don't pad content to increase time-on-page. unnecessary filler before the actual answer increases the odds someone bounces back to search results out of frustration, which is the pattern that plausibly hurts you.
- answer the actual query clearly and early — see put the answer above the fold — so a visitor gets what they came for and either leaves satisfied or keeps reading because they're genuinely interested, not because they're hunting for the answer you buried.
- match the page to the intent behind the query. a mismatch between what someone expected and what they found is the most common cause of pogo-sticking, more than anything about page length or design.
the takeaway
a visitor spending twenty seconds on your page and leaving happy because they got a clear answer isn't a ranking problem. a visitor bouncing back to search results to try someone else, over and over, plausibly is — but that's a satisfaction problem, not a time-on-page problem. optimize for answering the question well, not for keeping the clock running — which, in practice, usually comes down to clarity over a more polished but vaguer page.
nanushi builds pages designed around answering the actual query fast and clearly, rather than chasing vanity engagement metrics that don't hold up under scrutiny.