Engagement minutes are a measure of how much time an audience actively spends with content, counted only while there are signs that someone is present and paying attention. Scrolling, mouse movement or a keystroke keeps the clock running; an abandoned tab stops it within seconds. The figure is reported as a total across a page or site, and as an average per visitor or page. It exists because the older web metrics, pageviews and "time on site", counted arrivals and gaps between clicks rather than reading, and publishers wanted a number that rewarded content people actually stayed with.

There is no single industry definition. The phrase is used by individual analytics vendors, each with its own counting rules. Chartbeat calls its version engaged time, Parse.ly reports engaged minutes, Upworthy coined attention minutes, and the content marketing platform ScribbleLive labels its metric "Engagement Minutes". Google Analytics 4 reports a close cousin as user engagement. The Media Rating Council (MRC), the US body that audits media measurement, has accredited individual vendors' time metrics but has not published a standard for engaged time on content.

How the minutes are counted

Chartbeat publishes the most detailed rules. Its help centre lists seven browser events that count as engagement: load, focus, scroll, resize, mousemove, mousedown and keydown. Each event keeps the user "engaged" for five more seconds, and the windows overlap rather than add, so ten mouse movements within two seconds earn five seconds of credit, not fifty. The clock stops when the reader switches tabs or another application takes focus. While a reader is engaged, the script reports every 15 seconds. Chartbeat states that "we do not count video playing time toward our engaged time metrics", on the reasoning that a playing video does not prove anyone is watching.

Parse.ly, now owned by Automattic, uses a similar heartbeat. According to its documentation, "after five seconds of inactivity, the heartbeat no longer considers the visitor engaged, and time stops tracking". It treats scrolling, clicking and video playback as activity, but only if a site's developers set a PARSELY.videoPlaying value so the script knows a video is running. The two vendors therefore disagree on the most basic question for a video-heavy page: whether watching counts.

Google Analytics 4 measures from the browser's side. Google defines user engagement as "the amount of time someone spends with your web page in focus or app screen in the foreground", sent in milliseconds through the engagement_time_msec parameter. A session becomes an engaged session after 10 seconds by default, a timer site owners can raise to a maximum of 60 seconds, as PPC Land's guide to GA4 session timeouts set out in May 2024.

A worked example shows how far the totals can diverge. A reader opens a 1,200-word article, reads for three minutes, watches a two-minute embedded video without touching the mouse, then leaves the tab open in front of them for four minutes while on the phone. Chartbeat would record roughly three minutes, since video time is excluded and the idle stretch ends five seconds after the last event. Parse.ly would record about five minutes if the video flag is set. GA4 could record nine. ScribbleLive adds a further wrinkle: its documentation says "seconds are rounded up to the next minute", which inflates totals built from many short visits.

The metric sits on the sell side. Editors use it to judge which stories hold readers, and commercial teams use it to describe audiences to advertisers; Parse.ly calls it "a measure of 'satisfaction' with content". Chartbeat's active exposure time applies the same logic to the seconds an ad is viewable in front of an active reader.

Origin and evolution

Chartbeat was the first to make engaged time a headline number. On April 16, 2012, alongside a $9.5 million Series B round led by Draper Fisher Jurvetson, it redesigned its dashboard around total engaged time, average engagement per user and per page, according to TechCrunch. Tony Haile, its chief executive, became the concept's main advocate.

Video moved in the same direction that year. On October 12, 2012, YouTube changed its search ranking to "reward engaging videos that keep viewers watching", according to TechCrunch.

The phrase that circulated most widely came from a publisher. Upworthy announced in February 2014 that it would steer the company by attention minutes and released the code that June. "We actually use attention minutes as a core company goal," Ed Urgola, its head of marketing, told the Columbia Journalism Review on June 23, 2014, adding that such metrics were "hard to ignore and even harder to manipulate". In the same piece Clare Carr of Parse.ly said the company had begun offering attention minutes to all clients.

Accreditation and commercialisation followed. Chartbeat announced on September 29, 2014 that the MRC had accredited 21 of its metrics, including viewability and active exposure time. A Digital Content Next survey reported in October 2014 that 80% of premium publishers wanted to sell advertising on time spent. The Financial Times rolled out cost per hour (CPH) pricing on May 18, 2015, charging only when an ad was seen for more than five seconds of active use, with Chartbeat measuring. Dominic Good, the FT's ad sales director, said that "while CPM values every impression the same, CPH uses time to measure value", according to Fortune. The Economist followed on November 4, 2015, using Chartbeat and Moat, with a 30-second cap per impression and the same five-second floor. "Anything below five seconds we don't class as meaningful attention," Ashwin Sridhar, its global head of digital products revenue, told Digiday.

Google's own definitions shifted later. Google Analytics 4 replaced Universal Analytics' "time on site", which could not measure the final page of a visit, with focus-based engagement time, and Google set July 1, 2023 as the sunset date for Universal Analytics.

Why it matters for marketers

For publishers, engagement minutes are a way to sell quality rather than volume. A page that draws many visits but little reading looks strong on pageviews and weak on engaged time, which is why Carr described a high-pageview, low-attention story as a possible sign of a misleading headline. GA4 places average engagement time per session next to conversions in its reports; the landing page report Google introduced in December 2022 shows both for each entry page.

The metric has gained weight as search referrals fall. Chartbeat data reported by Nieman Lab on March 18, 2026 showed Google Search pageviews to publishers down 34% between December 2024 and December 2025, while chatbot referrals remained under 1% of pageviews. The same report found news sites receive the lowest engagement per article from AI platforms. With fewer arrivals, how long existing readers stay has become a larger part of a publisher's case to advertisers. Audio uses the same logic: Podstock reported in May 2026 that a Spotify stream generates about 1.5 times more listening time than a YouTube view of the same podcast episode.

Limitations and disputes

The central weakness is the lack of a common definition. A Digital Content Next report cited by Digiday in October 2014 found 95% of publishers used more than one vendor to measure time spent, and the rules have not converged since. The video split between Chartbeat and Parse.ly, GA4's focus-based approach and ScribbleLive's rounding mean two "engagement minute" figures can describe the same visit very differently.

Activity is also a proxy for attention, not proof of it. A mouse can move while a reader looks elsewhere, and a reader can study a chart without touching anything. The five-second idle window is a convention chosen by vendors, not a finding about human attention. Parse.ly concedes that "a long engaged time could indicate some sort of visitor confusion", so more minutes are not always better.

Critics have questioned whether time predicts advertising outcomes. Edward Kim, chief executive of SimpleReach, told Digiday in 2014 that "it still remains to be seen that time is the best indicator for success". Industry bodies kept their distance: Sherrill Mane of the Interactive Advertising Bureau (IAB) said in 2014 that the IAB "has not gone to MRC with the notion that attention measurement merits a task force", according to MarTech.

Watch time or minutes viewed. These count playback, not activity. YouTube's 8,000 watch-hour threshold for advertising revenue, set on August 10, 2026, and the Nielsen Gauge's share of TV minutes both credit a playing screen whether or not anyone interacts.

Average minute rating. A television audience estimate: the average number of people watching across each minute of a programme. It counts people per minute, not minutes per person.

Attention seconds, APM and cumulative attention. Attention metrics estimate whether eyes were on a specific ad, usually from eye-tracking models, in seconds. Engagement minutes describe time with content, not with ads.

Engagement minutes in webinars. ON24's glossary uses the same phrase for the total engaged minutes of all leads in an account, counting polls, questions, downloads and viewing. That is a business-to-business lead-scoring sense unrelated to publisher audiences.

Recent developments

The closest thing to an industry rulebook now covers ads rather than content. The MRC and IAB finalised attention measurement guidelines on November 12, 2025, allowing attentive time to be accumulated on a non-continuous basis within a session but leaving content engagement time outside their scope. TVision reported in August 2026 that only nine of more than 200 daily viewing minutes were attentive ad time. As of October 2026, no audited, cross-vendor definition of an engagement minute exists.

Timeline

  • 2009 - Parse.ly is founded by Sachin Kamdar and Andrew Montalenti.
  • April 16, 2012 - Chartbeat relaunches its dashboard around engaged time with a $9.5 million Series B.
  • October 12, 2012 - YouTube ranks search results by watch time, a day after adding watch-time data to YouTube Analytics.
  • February 2014 - Upworthy announces attention minutes as its core metric.
  • June 23, 2014 - Upworthy's attention minutes code release is reported; Parse.ly offers attention minutes to all clients.
  • September 29, 2014 - Chartbeat announces MRC accreditation for 21 metrics, including active exposure time.
  • October 2014 - Digital Content Next reports that 80% of premium publishers want to sell on time spent.
  • May 18, 2015 - The Financial Times rolls out cost-per-hour advertising with a five-second active-use floor.
  • November 4, 2015 - The Economist adopts cost-per-hour selling with a 30-second cap per impression.
  • February 8, 2021 - Automattic acquires Parse.ly.
  • July 1, 2023 - Universal Analytics stops processing data; GA4's engagement time replaces time on site.
  • November 12, 2025 - MRC and IAB finalise attention measurement guidelines for ads.
  • March 18, 2026 - Chartbeat data shows Google Search pageviews to publishers down 34% year on year.
  • August 10, 2026 - YouTube doubles its long-form watch-hour threshold to 8,000.

Summary

Who: Content analytics vendors such as Chartbeat, Parse.ly and ScribbleLive, Google through GA4, and the publishers, including the Financial Times and The Economist, that report or sell on engaged time.

What: Engagement minutes are the total or average time people actively spend with content, counted only while browser signals such as scrolling, mouse movement or keystrokes, or page focus, show that someone is present.

When: Chartbeat made engaged time a headline metric in April 2012; Upworthy popularised attention minutes in 2014; time-based ad selling followed in 2015; GA4's engagement time became Google's default in July 2023.

Where: In publisher and website analytics dashboards, newsroom tools, media kits and time-based ad products, mostly on the open web.

Why: Pageviews and time on site rewarded clicks rather than reading, so publishers wanted a measure of sustained use; the metric remains vendor-defined, with no MRC standard and material differences in how video and idle time are counted.