The Attention Landscape · established evidence

The Attention Fragmentation Index: What Multi-Platform Media Diets Actually Look Like in 2026

Last reviewed 2026-07-20. Written by Chandranshu Kumar, Founder, Raveneye Global. · 9 min read

A typical adult is not loyal to one platform; GWI's tracking of 250,000-plus respondents across 50-plus countries finds the average person active on 6.5 to 6.75 distinct platforms a month. At the same time, the total daily budget spent on social platforms has been shrinking, not growing, since it peaked in 2022, down to roughly 2 hours and 20 minutes a day among adults in the developed world by late 2024, about 10 percent below the peak. Put those two established findings together and a specific, testable claim follows: attention is not concentrating on fewer surfaces, it is spreading thinner across more of them at once. A visibility plan built around a single channel, however well executed, is measuring and defending a shrinking slice of a fragmenting whole. This piece lays out the evidence for that fragmentation and what it implies for how a business should actually measure where its buyers spend their attention.

The number that defines "fragmented": 6.5 to 6.75 platforms a month

GWI's tracking panel, which surveys more than 250,000 respondents across more than 50 countries, is one of the largest and most methodologically disclosed sources on multi-platform behavior available. Its most recent findings put the average person's monthly platform footprint at 6.5 to 6.75 distinct platforms, alongside a combined weekly social-plus-video time figure of roughly 18 hours and 36 minutes.

That figure is worth sitting with. It does not describe a person who picks a primary platform and occasionally checks a second one. It describes someone whose daily attention is routinely distributed across six or seven separate destinations, each with its own algorithm, its own content format, and its own relationship to how that person discovers and trusts a business.

The time budget peaked in 2022, and has been falling since

The second half of the fragmentation story is less intuitive and more important. Total time spent on social platforms is not still climbing. GWI's panel shows social-media time peaked in 2022 and has declined since; by the end of 2024, adults 16 and older in the developed world spent roughly 2 hours and 20 minutes a day on social platforms, about 10 percent below the 2022 peak.

That is the correction to a common assumption. The story most owners carry in their heads is "people spend more time online every year, on more platforms." The first half is now false. Total social time is shrinking. The platform count that time gets spread across is not shrinking with it. The two trends together, a falling time budget divided across a stable-to-growing number of destinations, is the precise mechanism of fragmentation: less attention, cut into more, smaller pieces.

Where the shrinking budget is being reallocated

Fragmentation inside the social-time budget is not evenly spread either. Short-form video has absorbed a disproportionate share of whatever attention remains. Platform-disclosed figures show Reels accounting for roughly 46 to 50 percent of all time spent on Instagram, and YouTube Shorts logging around 200 billion views a day, both established, company-reported figures. eMarketer's forecast, by contrast, has average US TikTok time falling to about 52 minutes a day in 2025, down roughly 6.9 percent year over year, even as broader industry trackers cite global in-app TikTok time closer to 95 minutes a day, a sign that US-market saturation and direct competition from Reels and Shorts is capping one platform even as the format itself keeps growing overall.

A cross-platform aggregate figure often repeated in industry content, that short-form video consumption rose roughly 75 percent with users spending 80-plus minutes a day across the category, appears consistently in secondary sources but was not traced to a single, clearly disclosed primary methodology for this piece. It should be treated as directional only, not cited as a precise industry figure, until a named primary source is located.

A non-vendor check: what a statutory regulator sees

Nearly every figure in this piece comes from an industry panel with a commercial interest in the category it measures, GWI, eMarketer, and platform self-disclosure among them. That is disclosed above and should be weighed accordingly. A useful, structurally different check exists in Ofcom, the UK's communications regulator, whose Online Nation and Media Nations reports are built on a large probability-based panel and passive metering rather than a vendor-recruited sample, with methodology published alongside each release.

Ofcom's device- and platform-time-split data corroborates the same directional finding independently sourced above, that a typical user's daily media consumption is now routinely split across multiple concurrent platforms and devices, a phenomenon usually described as multi-screening. This piece does not cite a specific Ofcom percentage as canon here, since no single figure from that source was independently verified for this article, but the existence of a non-vendor, statutory source pointing the same direction as the commercial panels above meaningfully strengthens the underlying claim.

Why "just do SEO" already under-serves this buyer

A business that concentrates its entire visibility effort on one surface, classic search being the most common single-channel default, is optimizing for roughly one-sixth to one-seventh of where a typical buyer's monthly attention actually lands, based on the GWI platform-count figure above. That is true even before accounting for the AI-answer layer, which sits outside GWI's traditional social-platform count entirely and represents a further, separate fragmentation of the same shrinking time budget.

This is not an argument for being everywhere at once, which is a separate and usually worse mistake: spreading a fixed effort across seven surfaces thinly produces weak presence on all of them. It is an argument for measuring where a specific business's buyers actually concentrate their fragmented attention, rather than assuming it mirrors the one surface a business has historically invested in.

The measurement problem fragmentation creates

Fragmentation has a second-order consequence that most reporting tools were never built to handle: a business's own analytics only ever show the surface that analytics tool is watching. Google Search Console shows search. A social scheduler shows the two or three platforms connected to it. None of them show the other four or five platforms a typical buyer also uses that month, and none of them show the AI-answer layer at all, since no assistant currently publishes citation logs to the businesses it names.

The practical effect is that a business's internal reporting almost always overstates how well it understands its own visibility, because it is reading one or two channels out of six or seven and treating that partial read as the whole picture. Fixing that requires deliberately sampling across the surfaces a buyer actually inhabits, not just the ones a business happens to already be instrumented for.

What a careful fragmentation read looks like

A careful attempt to map this for a specific business has to do three things the GWI-level aggregate data cannot do on its own. It has to establish which of the six to seven platforms a typical monthly user touches actually carry that business's specific buyer population, since GWI's average is a global blend across every demographic and vertical, not a prediction for any one business. It has to separately account for the AI-answer layer, which behaves nothing like a traditional social platform and is invisible to almost every existing analytics stack. And it has to be repeated on a cadence, since the underlying mix keeps moving, as the short-form-video reallocation above already shows within a single two-year window.

None of that is a reason to distrust the macro finding. Social time shrinking since 2022 while platform count per person stays at 6.5-plus is well evidenced, from a large, methodologically disclosed panel, corroborated directionally by a non-vendor regulator. It is a reason to treat the macro number as the starting frame, not the finished answer, for any one business's actual visibility map.

The evidence

Key findings, with their sources

The average person is active on 6.5 to 6.75 distinct platforms a month, with combined weekly social-plus-video time of roughly 18 hours 36 minutes.

established GWI, "Time spent online declines" press release; DataReportal, "Digital 2026: Global Overview Report."
Social-media time peaked in 2022 and has since fallen; by end-2024 adults 16-plus in the developed world spent roughly 2h20m/day on social platforms, about 10% below the 2022 peak.

established GWI panel tracking, 250,000+ respondents across 50+ countries.
Reels account for roughly 46 to 50 percent of all time spent on Instagram; YouTube Shorts logs around 200 billion views per day.

established Instagram/Meta and YouTube, platform-disclosed figures.
eMarketer forecasts average US TikTok time falling to about 52 minutes a day in 2025, down roughly 6.9% year over year, even as other trackers cite global in-app time near 95 minutes a day.

established eMarketer, "US TikTok Usage and Time Spent 2025."
A commonly repeated claim that short-form video consumption rose 75% with 80-plus minutes/day of use has no single clearly disclosed primary source and should be treated as directional only.

contested Cross-referenced against multiple secondary aggregator sources; no named primary methodology located.

Reference

Glossary

Attention fragmentation: The distribution of a shrinking total attention budget across a stable or growing number of distinct platforms and surfaces, rather than concentration on one or two dominant channels.
Platform footprint: The number of distinct platforms an individual actively uses within a given period, most commonly measured monthly. GWI's panel puts the current average at 6.5 to 6.75.
Multi-screening: The behavior of using more than one device or platform concurrently, a phenomenon documented independently by non-vendor sources such as Ofcom.
Non-vendor source: A data source without a commercial stake in the category it measures, such as a statutory regulator (Ofcom) or a government statistical agency, used here as a corroborating check on vendor panel data.

Straight answers

Frequently asked questions

How many platforms does the average person actually use?

GWI's panel of 250,000-plus respondents across more than 50 countries finds the average person active on 6.5 to 6.75 distinct platforms a month, with roughly 18 hours 36 minutes of combined weekly social-plus-video time.

Is social media time still growing?

No. GWI's tracking shows social-media time peaked in 2022 and has been declining since, down to roughly 2 hours 20 minutes a day among adults in the developed world by late 2024, about 10% below the 2022 peak.

If people spend less time on social media, why does fragmentation matter more, not less?

Because the shrinking time budget is being divided across a stable-to-growing number of platforms, not concentrated onto fewer of them. Less total attention, cut into more pieces, is the specific mechanism that makes any single-channel strategy under-serve a modern buyer.

Does this include AI assistants like ChatGPT?

No. GWI's platform-count figure measures traditional social and video platforms. The AI-answer layer is a separate, additional fragmentation of the same shrinking attention budget, and it is invisible to almost every standard analytics tool a business currently uses.

What should a business actually do with a macro fragmentation statistic?

Treat it as the frame, not the answer. The GWI average is a global blend across every demographic and vertical. The useful next step is mapping which of the six-plus platforms a specific business's own buyer population actually concentrates on, and repeating that read on a cadence, since the mix keeps shifting.

Provenance

Sources

GWI, "Time spent online declines" press release, gwi.com/press-releases/time-spent-online (established)gwi.com
DataReportal / Meltwater / We Are Social, "Digital 2026: Global Overview Report," datareportal.com (established)datareportal.com
eMarketer, "US TikTok Usage and Time Spent 2025" (established)emarketer.com
Instagram/Meta and YouTube, platform-disclosed usage figures on Reels and Shorts (established)
Ofcom, "Online Nation 2025" and "Media Nations UK 2025," ofcom.org.uk (established, non-vendor corroboration)ofcom.org.uk
Secondary aggregator figures on cross-platform short-form video totals, no primary source verified (contested, flagged directional only)

Every figure above is attributed to a real, dated source and tagged with its evidence tier. Where a claim could not be verified to a primary source, it is not stated as fact.