Recent changes observed in how videos surface across YouTube’s homepage and search results are reshaping visibility for creators, with analysts pointing to increased promotion of short-form content and reduced placement for long-form videos.

The shifts have been documented by Marcus Jones, a YouTube consultant who analyzed performance data from more than 500 channels across multiple creator size tiers. According to Jones’ analysis, creators who previously averaged around 150,000 views per long-form upload are now seeing closer to 60,000, even as their overall publishing cadence remains unchanged.

Separate data from analytics firm Velio indicates that the top 100 YouTubers collectively experienced a roughly 50% decline in views over a single week during the same period Jones identified the distribution changes.

Homepage Real Estate for Long-Form Content Continues to Shrink

One of the clearest signals identified in Jones’s review is a structural change to YouTube’s homepage layout.

In 2024, the homepage typically surfaced two rows of five long-form videos, giving creators roughly ten opportunities to compete for prominent placement. By early 2025, that layout shifted to two rows of three long-form videos. Current homepage configurations observed across multiple regions now show just one row of two to three long-form videos, with Shorts occupying most remaining screen space.

“Instead of fighting for one out of ten available slots on the homepage like back in 2024, you’re now fighting for one out of two to three available spots,” Jones said in his analysis.

The same pattern appears across devices. On mobile, Shorts blocks increasingly dominate the first screen users encounter when opening the app, while search results surface a higher volume of short-form content ahead of longer videos.

Traffic Source Data Shows Sustained Shorts Promotion

Jones’s analysis of traffic source data suggests Shorts visibility from homepage and search placements roughly doubled compared with 2024 levels. The increase became particularly pronounced in April and May 2025 before stabilizing at levels well above prior baselines.

Jones notes that these distribution changes do not necessarily indicate a reduction in total long-form views across the platform as a whole. Instead, his data suggests those views are being concentrated among a smaller group of channels, narrowing the number of creators capturing consistent long-form traffic.

The findings reflect trends observed across Jones’s dataset and are not positioned as platform-wide confirmation.

Older Shorts See Steep Declines in Distribution

In a separate analysis, Mario Joos, retention director who’s worked with creators including MrBeast and the Stokes Twins, documented declines affecting previously published Shorts.

Reviewing data across channels generating between 100 million and 1 billion monthly views, Joos identified a consistent drop-off in performance for Shorts published more than 30 days earlier.

“After digging through data across multiple channels, including ones that pull 100 million to 1 billion views per month, we finally found the pattern,” Joos wrote in a LinkedIn post. “The issue wasn’t with new content. The issue was with older content, specifically Shorts that were published more than one month ago.”

According to Joos, the shift appears to have begun around mid-September 2025, with distribution prioritizing Shorts published within the past 28 to 30 days while reducing promotion of older, previously evergreen videos.

“Older Shorts, even strong evergreen ones, are suddenly getting flattened,” he wrote.

Image source: Mario Joos

AI-Generated Content Adds Competitive Pressure Inside Recommendations

Separate research suggests another factor intensifying competition within YouTube’s recommendation system: the growth of AI-generated content.

A study from video-editing company Kapwing examined 15,000 channels and identified 278 channels composed entirely of AI-generated videos. Collectively, those channels have accumulated more than 63 billion views and 221 million subscribers, according to the firm’s analysis.

Kapwing also created a new YouTube account and reviewed its first 500 recommended videos, finding that 104 were AI-generated. The company estimated that these channels generate approximately $117 million in annual revenue.

The most-viewed channel in the study, “Bandar Apna Dost,” based in India, has accumulated 2.4 billion views, while other high-performing AI content channels include “Pouty Frenchie” in Singapore and “Cuentos Fascinantes” in the United States.

A YouTube spokesperson told The Guardian that the platform terminated one AI-generated channel identified in Kapwing’s report and removed others from monetization programs, citing enforcement of community guidelines.

While the AI findings are distinct from Shorts distribution changes, analysts say they contribute to broader competition for recommendation space, particularly as homepage and search real estate becomes more constrained.

Sustainability Questions for Creator Businesses

Jones said the distribution trends raise longer-term questions about creator sustainability, particularly for businesses built on long-form content. Short-form videos typically generate lower advertising revenue and command smaller sponsorship fees, he noted, while also requiring higher output volume.

“If YouTube becomes a platform where more and more creators have to work three times as hard for 50% of what they’re currently getting, the chances of those creators quitting increases significantly,” Jones said.

Joos’s post also raises the same concerns, warning that reduced distribution for evergreen content shifts incentives away from long-term value creation.

“This change pushes everyone toward quantity over quality and makes it harder to build sustainable revenue or long-term video value,” he wrote.

Jones said he has communicated his findings to YouTube representatives, whom he described as “genuinely good people” focused on optimizing for viewer behavior. YouTube has not publicly addressed the specific algorithmic or distribution changes documented in the analyses.

Checkout Our Latest Podcast