How Instagram Reels Distribution Works: What Meta Has Publicly Shared
Instagram uses multiple ranking systems — not one algorithm — to distribute Reels. Publicly confirmed top signals include watch time, likes per reach, and sends per reach, according to Instagram head Adam Mosseri in January 2025. Hanami Social's data across 430M+ client views aligns with these official statements and adds practical context for businesses producing short-form video.
Matt Hannan
What Instagram Has Officially Said About Reels Ranking
Instagram does not use a single algorithm. The platform uses multiple ranking systems tailored to different surfaces — Feed, Stories, Explore, and Reels each have their own approach (official ranking blog). For Reels specifically, the ranking system evaluates four categories of signals: your activity, your history with the poster, information about the reel itself, and information about the poster. The primary predictions the system makes are whether a viewer will reshare the Reel, watch it to completion, like it, or visit the audio page.
Having worked directly on these publishing systems at Meta, I can add context: the infrastructure that evaluates these signals operates at extraordinary scale and speed. Every piece of content is scored against hundreds of features within milliseconds of being eligible for distribution. Understanding which signals the system weighs most heavily gives creators and businesses a meaningful advantage.
What Are the Confirmed Top Ranking Signals for Reels?
Instagram head Adam Mosseri stated in January 2025 that the top three ranking factors for Reels are watch time, likes per reach, and sends per reach. This is significant because it clarifies the relative weight of different engagement types. For connected reach (people who already follow you), likes are slightly more important than sends. For unconnected reach (the Explore and Reels tabs), sends are more important than likes.
This distinction matters for businesses. If your goal is to reach people who do not yet follow you — which is the primary growth mechanism — optimizing for shareability is more impactful than optimizing for likes alone.
| Ranking Signal | Source | Context |
|---|---|---|
| Watch time | Mosseri, January 2025 | Top 3 signal for Reels distribution |
| Likes per reach | Mosseri, January 2025 | Top 3 signal; slightly more important for connected reach |
| Sends per reach | Mosseri, January 2025 | Top 3 signal; more important for unconnected reach |
| Reshares | Instagram official blog | Listed as primary Reels prediction |
| Completion rate | Instagram official blog | Listed as primary Reels prediction |
| Audio page visits | Instagram official blog | Listed as primary Reels prediction |
| Low resolution | Instagram official blog | Quality filter that reduces visibility |
| Watermarks | Instagram official blog, Mosseri | Content with watermarks gets deprioritized |
| Excessive text overlay | Instagram official blog | Quality filter that reduces visibility |
| Borders/muting | Instagram official blog | Quality filters that reduce visibility |
Every entry in this table comes from either Instagram’s official ranking explanation or Mosseri’s public statements. We have not assigned numerical weights because Instagram has not published them.
How Does Watch Time Affect Reel Distribution?
Watch time is the single most cited signal across every official Instagram statement about Reels ranking. Across our client campaigns totaling 430M+ views, Reels that hold attention through the full duration consistently outperform those that generate quick engagement but lose viewers early. This aligns with what Instagram has publicly stated: completion rate is one of the primary predictions the ranking system makes for Reels.
Mosseri also confirmed in January 2025 that videos over 3 minutes will not get recommended to non-followers. This establishes a hard ceiling on content length for anyone pursuing growth through unconnected reach.
The practical implication is straightforward. The first few seconds of a Reel determine whether someone keeps watching. The opening hook — what happens before a viewer decides to swipe away — has an outsized impact on overall watch time. This is why Hanami Social maintains a database of 1,460+ viral hook templates, each catalogued by format, topic, and observed performance patterns.
Do Likes Actually Matter for Reels?
Yes — likes are one of the top three ranking signals, according to Mosseri’s January 2025 statement. For connected reach (distribution to your existing followers), likes are slightly more important than sends. For unconnected reach (distribution to non-followers via Explore and the Reels tab), sends carry more weight than likes.
This means likes are far from irrelevant. They are one of the three most important signals Instagram uses to decide how widely to distribute a Reel. The common advice to “ignore likes and focus on shares” oversimplifies what Instagram has actually said. Both matter — the question is which audience you are trying to reach.
Reels that generate both high likes and high sends tend to perform best overall, because they satisfy the ranking criteria for both connected and unconnected distribution. The YukiHomes campaign — where 10 Reels generated 900+ leads in 3.5 weeks — produced content that performed well on both metrics, contributing to broad distribution across both follower and non-follower audiences.
What Content Gets Penalized by Instagram?
Instagram’s official ranking blog explicitly lists quality filters that reduce a Reel’s visibility. These are not speculation — they are published by Instagram:
- Low resolution video — Blurry or heavily compressed content gets deprioritized
- Watermarked content — Mosseri confirmed that content with watermarks (including TikTok watermarks) gets deprioritized
- Videos with borders — Content that does not fill the full screen
- Muted videos — Reels uploaded without audio
- Excessive text overlay — Too much on-screen text covering the visual content
- Reposted content — Content that has been previously posted or is not original
Additionally, Mosseri’s December 31, 2025 memo indicated that Instagram would be prioritizing “raw, real human content” over AI-generated material in 2026. While the exact mechanisms for detecting and deprioritizing AI content have not been detailed publicly, the directional signal is clear: original, human-created content is what the platform wants to promote.
How Does Explore Distribution Differ from Feed?
Popularity matters “much more” in Explore than it does in Feed or Stories (Instagram official ranking blog). This means the bar for reaching non-followers through Explore is higher — content needs to demonstrate broader appeal, not just resonate with your existing audience.
This is where the initial engagement window matters most. A Reel that generates strong early signals from your followers has a better chance of crossing into Explore distribution, where it can reach entirely new audiences. This is the mechanism behind viral growth on the platform.
This pattern shows up consistently in our client work. For example, when we produced content for a rock band client, the first Hanami Social Reel achieved 111,815 views and 4,547 interactions, reaching 47,720 accounts. Over a two-month comparison period, the account saw +122% views (3.06M to 6.81M), +163% interactions (49.6K to 130.1K), and +138% non-follower reach (185K to 441K). The growth in non-follower reach — which is primarily driven by Explore and Reels tab distribution — outpaced overall view growth, consistent with Instagram’s stated emphasis on popularity signals for Explore.
What Does This Mean for Businesses?
The publicly available information from Instagram points to a clear set of priorities for businesses producing Reels:
Optimize for watch time first. Every official statement positions watch time or completion rate as a primary ranking signal. This means the hook — the opening seconds that determine whether someone keeps watching — deserves disproportionate creative investment.
Create content people want to send to others. Sends per reach is a top-three signal, especially for reaching non-followers. Content that is useful, surprising, or identity-reinforcing tends to get shared. Data-driven content (specific numbers, comparisons, reveals) consistently generates higher send rates than purely entertaining content, though both have a place in a content strategy.
Avoid the published quality filters. There is no reason to lose distribution to preventable issues like watermarks, low resolution, or borders. These are explicitly confirmed penalties.
Keep Reels under 3 minutes. Mosseri confirmed this hard limit for recommendation eligibility. Most high-performing Reels for business accounts are significantly shorter — typically under 60 seconds — but the ceiling is worth knowing.
Pair distribution with lead capture. High view counts alone do not generate revenue. Hanami Social’s HanamiDM system converts Reels engagement into captured leads by automating DM conversations triggered by keyword comments. Across all clients, this approach has captured 15K+ automated leads, turning algorithmic distribution into measurable business outcomes.
Turning Distribution Into Results: Real Client Data
Understanding how distribution works is necessary but not sufficient. The question that matters for businesses is: does this knowledge translate into results? Here is what Hanami Social has achieved by applying these principles:
| Metric | Result |
|---|---|
| Total views across clients | 430M+ |
| Largest single Reel | 6.9M views |
| Follow actions generated | ~500K+ |
| Automated leads captured | 15K+ |
| YukiHomes: leads from 10 Reels in 3.5 weeks | 900+ |
| YukiHomes: estimated revenue value | ~$150K at 3.5% conversion |
| YukiHomes: ROI on content production | ~30x |
| English course client: leads from 3 videos in 1 week | 170 |
These results come from treating each Reel as a data point in a larger system, not as an isolated creative bet. When you understand what the platform has publicly said about how it distributes content, you can design content that aligns with those signals — and then measure whether your execution actually performs.
Key Takeaways
- Instagram uses multiple ranking systems, not one algorithm — Reels has its own distinct ranking approach
- The top three Reels ranking signals are watch time, likes per reach, and sends per reach (Mosseri, January 2025)
- Likes are one of the top three signals, not a weak metric — they matter more for connected reach, while sends matter more for unconnected reach
- Quality filters (watermarks, low resolution, borders, excessive text) are explicitly confirmed to reduce distribution
- Videos over 3 minutes will not get recommended to non-followers
- Instagram is prioritizing raw, human content over AI-generated material in 2026
Matt Hannan is the founder of Hanami Social and a former Meta Senior Software Engineer who worked directly on the Instagram Reels publishing system. He also holds a US Patent Application for video reformatting technology. Hanami Social helps businesses in Japan and the United States grow through data-driven Instagram Reels production and automated lead capture. Book a free strategy call.
Related Questions
- Q: How does the Instagram Reels algorithm work in 2026?
- Q: What makes an Instagram Reel go viral?
- Q: How to grow on Instagram Reels?
- Q: What is the best strategy for Instagram Reels growth?
- Q: How does Instagram decide which Reels to show?