Who Owns Dwell Time, the Search Engine or Your Analytics

What is dwell time?

Search engines measure dwell time as the interval between a searcher clicking a result and returning to the search results page, counted on the engine's side, not by your analytics script. The engine serves the results page, records the outbound click, and records the searcher landing back on it, so every event sits in its own logs. Your tag fires once in the middle of that sequence and sees neither the click that started it nor the return that ended it.

dwell time = timestamp of return to the SERP - timestamp of the result click

A searcher clicks your pricing page from Google at 14:02:10 and presses back at 14:03:35. Dwell time for that visit is 85 seconds. Your analytics recorded one pageview and, under the Universal Analytics model, zero seconds of time on page, because time on page is the gap between two hits and there was no second hit. One visit, 85 seconds on the engine's side of the line and 0 on yours.

How is dwell time different from time on page and session duration?

The three metrics start and stop different clocks, and only one is measured by someone other than you. Dwell time is bounded by two events on the search results page, time on page by two hits from your own tag, and session duration by a session's first and last hit.

That last column is the one people skip. A vendor can sell you time on page and session duration because your script produced both. Nobody outside Google or Bing can sell you dwell time, because your script was not running on the results page.

For the session clock and the zero-second last pageview, see how average session duration is calculated and why time on page in Google Analytics is a flawed metric. The boundary that decides where the second clock stops is covered in what counts as a session, and the separate bounce and exit pair is worked through in bounce rate vs exit rate.

Who can measure dwell time?

Only the search engine that served the results page can measure dwell time. Both timestamps in the formula belong to the SERP, and a first-party script cannot see a page it does not run on. Your tag can record when the tab was hidden through the pagehide event, but a hidden tab does not say whether the person went back to Google, closed it, or answered the door.

Treat any product advertising a dwell time column as reporting something else under a borrowed name. Ask which two events bound the number. If both came from your own script, you are looking at time on page with a new label.

Does Google use dwell time as a ranking factor?

Google does not name dwell time as a ranking factor in any of its public Search documentation. Google's "A guide to Google Search ranking systems" names its documented systems, including BERT, RankBrain, neural matching, link analysis systems and PageRank, and dwell time appears nowhere in that list (checked September 9, 2026). The SEO starter guide's section "Things we believe you shouldn't focus on" does not mention dwell time, time on site or bounce rate either, and the "Creating helpful, reliable, people-first content" page does not use the term at all.

What Google documents is coarser. Its How Search Works page on ranking results states: "We also use aggregated and anonymized interaction data to assess whether search results are relevant to queries," and adds, "We transform that data into signals that help our machine-learned systems better estimate relevance" (checked September 9, 2026). That confirms aggregated interaction data feeds relevance estimation. It does not define dwell time or describe a per-page timer.

Quotes attributed to Google staff denying dwell time as a ranking signal circulate widely in SEO writing. None appear in Search Central documentation, so the honest description is that Google's position on dwell time is undocumented, not officially denied. Build your case on the aggregated interaction data sentence Google published, and label anything past it as inference.

What is pogo-sticking, and how is it different from dwell time?

Pogo-sticking is a searcher returning to the results page and clicking a different result for the same query. Dwell time is a duration; pogo-sticking is a sequence of clicks. A pogo-stick contains a short dwell time, but a short dwell time alone is not a pogo-stick, because the searcher might have closed the tab or read the one line they came for and stopped searching.

That last case matters for anyone reading short intervals as failure. A searcher who wanted your opening hours and got them in four seconds produced a four-second dwell time and a finished task. A searcher who could not find your pricing and clicked a competitor produced the same interval and an abandoned task. Identical duration, opposite outcome, which is why no engine publishes an ideal dwell time and why you should not adopt one.

What makes a searcher return to the results in a few seconds?

Searchers go back when the page does not answer the promise the search result made. The causes you can act on are a title describing something the page lacks, an answer buried below three screens of introduction, a first paint slow enough that nothing is readable, a consent wall over the content, and an interaction that does not respond when clicked.

Each leaves a fingerprint on your side of the line, even though the return itself does not. Scroll data shows whether people reached the answer, covered in what scroll depth measures. Repeated clicks on an element that does nothing show up in rage click analysis.

What should you track instead of dwell time?

Track the query-side numbers in Google Search Console and the behavior-side numbers on your own site, then read them together. Search Console gives you impressions, clicks, click-through rate and average position per query, and it does not report dwell time either. On your site, track scroll depth to the answer, completion of the step the page exists to drive, and error events in the first ten seconds.

For the qualitative half, watch what happened. Flowsery's AI session replay records every session, detects rage clicks, dead clicks, JavaScript errors and drop-offs automatically, groups matching sessions into one issue, and ranks issues by how many users hit them. That turns "searchers leave this page fast" into a named defect with the replay attached, which no dwell time number gives you.

Answers now appear on the results page itself, which changes what a click and a return mean. That shift is covered in SEO in the age of zero-click search.

Frequently Asked Questions

Is dwell time the same as bounce rate?

Dwell time and bounce rate count different things and belong to different parties. Dwell time is a duration the search engine measures between a result click and a return to the results page. Bounce rate is a ratio your analytics measures, dividing single-page sessions by the sessions that started on the page. A visit can be a bounce and still carry a long dwell time.

Can Google Analytics show dwell time?

Google Analytics cannot show dwell time, because it never sees the search results page. Its clocks start and stop when your tag fires, so it reports engagement time, time on page and session duration. Any dashboard labeling one of those as dwell time has renamed a first-party metric.

What is a good dwell time?

No search engine publishes a dwell time threshold, and no benchmark exists to check your pages against. A four-second visit that answered the query and a four-second visit that sent the searcher to a competitor produce the same number. Compare a page against its own completion rate instead: did people reach the answer, and did they act on it.

Does Google Search Console report dwell time?

Search Console does not include a dwell time metric. It reports impressions, clicks, click-through rate and average position per query, all measured on the results page before the click. Use it to find pages with high impressions and low click-through rate, then use replay to see what happens after the click.

Is a long dwell time always good?

A long interval before the return can mean careful reading or a person stuck in a broken form, and both look identical as a duration. Open the replays for long visits on a page that drives a conversion and check whether the extra minutes went into the flow or into fighting it.

How does dwell time relate to pogo-sticking?

Pogo-sticking is a short dwell time followed by a click on a different result for the same query. The duration is one half of the pattern, and the second click is the half that says the searcher was not satisfied. Neither event reaches your analytics, so diagnose the page through scroll depth, error events and replay.

Can the pagehide event measure dwell time?

The pagehide event only tells your script that the tab was hidden, not what the searcher did after that. A hidden tab does not say whether the person returned to the search results page, closed the browser, or stepped away for an unrelated reason. Dwell time needs a timestamp from the SERP itself, and pagehide only ever fires on your page.

Does Bing measure dwell time the same way Google does?

Bing runs its own results page, so the same logic applies: both timestamps sit in Bing's logs, not yours. The post's only claim about Bing is that it stands alongside Google as one of the two parties positioned to measure a real dwell time number. Nothing here documents how Bing calculates or uses that number beyond that.

How can I tell if a vendor's dwell time metric is real?

Ask which two events bound the number the vendor calls dwell time. If both events come from your own script, a first hit and a later hit, you are looking at time on page wearing a borrowed name. A genuine dwell time figure needs both timestamps from the search results page, and no tool besides the engine that served that page runs there.

What does Google's How Search Works page say about interaction data?

Google's How Search Works page on ranking results states that it uses "aggregated and anonymized interaction data to assess whether search results are relevant to queries" and turns that data into signals for its machine-learned relevance systems. That sentence confirms interaction data feeds ranking, but it does not define dwell time or describe a per-page timer. Anything past that sentence, including a claim that dwell time itself is a ranking factor, is inference rather than documentation.