What Is Pogo-Sticking?

The image is the pogo stick: the user hops onto a page and lands straight back on the list. In usability, Jared Spool used the term for users who jump up and down a site's hierarchy without finding what they want (talk "The Secret Lives of Links", recorded in November 2012). In SEO, Rand Fishkin covered it in Moz's Whiteboard Friday "Solving the Pogo-Stick Problem" on August 23, 2013. Nielsen Norman Group described the pattern in 2015 and noted that it is often discussed in connection with people returning to an external search engine.

The pattern has three steps: a click on a result, a quick return to the SERP, and a click on another result.

The causes usually cited sit on the landing page:

• The page doesn't match the search intent: someone looking for a price gets a guide, or the other way round.
• The answer is far down the page and not visible on arrival.
• The page loads slowly or is crowded with pop-ups and ads.
• The title or description in the result promises something the page doesn't deliver. Clickbait is the extreme case.

Nielsen Norman Group puts it as a mismatch between what the link promised and what the user finds on arrival.

One case looks identical and isn't a failure. People comparing prices, reviews or spec sheets open several results on purpose. John Mueller made this point in 2018 when asked about the pattern: users go back and forth for many reasons. From the outside, the two behaviors look the same.

Someone searches "how to clean an induction hob". The first result is a shop page for cleaning products with no instructions, and they are back within five seconds. The second result shows the steps at the top of the page, and they stay. For the first result that was pogo-sticking; for the second, a successful visit.

A different case: someone searches "cheap trail running shoes" and opens four shops one after another to compare them. The pattern is the same, but no page failed here, because comparing was the goal.

You can't measure it directly. Only Google sees the return to its results list, and it doesn't report it in any tool. What you can use are indirect signals:

• In Search Console, organic CTR per query. Search Console counts a click when a result sends the user outside Google Search; what the user does next isn't in the report.
• In Google Analytics 4, engagement rate and bounce rate for sessions from organic search. Google's help page says "The bounce rate is the opposite of the engagement rate". A session counts as engaged if, among other criteria, it lasts longer than 10 seconds.

A worked example with round numbers: a page gets 1,000 organic visits a month for one query, and 70% of the sessions are not engaged. That tells you many visits end quickly. It doesn't tell you how many of those visitors went back to Google and opened another result.

As a concept, pogo-sticking is useful. It describes, from the user's side, what goes wrong on a page that wins clicks and fails to hold them, and it pushes the analysis toward search intent rather than keywords. As a metric it is weak. Nobody outside Google can count it, and deliberate comparison shopping looks exactly like a disappointing page.

Source: Google Analytics Help, "Engagement rate and bounce rate"; Bing Webmaster Blog, August 2, 2011 (accessed 10.11.2026).

On whether Google uses pogo-sticking for ranking, two kinds of sources give different answers.

Google has said no. In a webmaster hangout in July 2018, John Mueller answered a question about this pattern: "We try not to use signals like that when it comes to search." There are many reasons for going back and forth, he said, so it is "really hard to refine" into a ranking factor. That is Search Engine Journal's transcript of July 16, 2018. In a Reddit Q&A on February 8, 2019, Gary Illyes went further and dismissed dwell time and CTR as baseless theories.

Against that stands a court ruling. In the antitrust case United States v. Google LLC, Judge Amit Mehta found on August 5, 2024, based on witness testimony including from Google employees, that the Navboost ranking system remembers which documents users clicked after entering a query and is trained on 13 months of user data (before 2017: 18 months). Google itself also writes on its How Search Works pages that it uses "aggregated and anonymized interaction data to assess whether search results are relevant to queries".

Both can be true at once. Google uses clicks in its ranking systems, and no primary source says whether a quick return to the results list counts as a signal of its own there. The internal documents leaked in May 2024 include fields such as "goodClicks" and "badClicks" in a Navboost module. On May 29, 2024, Google cautioned against assumptions based on "out-of-context, outdated, or incomplete information", and a field name says nothing about how the field is used.

• Search the query yourself and look at the page type at the top: guide, comparison, product page or shop.
• Check whether your page is that type, or whether the query needs a separate page.
• Put the answer where visitors see it on arrival, without scrolling.
• Read the title and description as a searcher sees them on the SERP and check that the page delivers what they promise.
• Check loading on mobile and remove anything that covers the content on arrival.

• Look in Search Console for queries with a good position and low CTR, and in GA4 for organic landing pages with low engagement. Those are your candidates.
• Judge each query on its own, because a comparison query produces back-and-forth even when the page is good.
• Improve the page for the people who visit it, not for a signal you can't measure.

• Treating GA4 bounce rate as if it were pogo-sticking.
• Claiming Google penalizes pogo-sticking: no primary source says so.
• Claiming the opposite, that Google uses no click data at all: the 2024 ruling says otherwise.
• Padding the text to keep visitors longer instead of giving them the answer sooner.

This entry was written in November 2026 with AI assistance. The facts come from the ruling in United States v. Google LLC, the Google Analytics and Search Console help pages, Search Engine Journal's transcript, and Internet Archive copies of Reddit and the Bing blog. It was reviewed before publication and contains no client data.