Last updated: November 18, 2026
Quick answer
A Reel reaches millions when a small test audience of mostly strangers watches it through and sends it to other people. Instagram head Adam Mosseri names watch time, likes per reach and sends per reach as the top ranking signals, with sends mattering slightly more for people who don't follow you. A strong first three seconds, an original idea and a reason to share matter more than follower count, hashtags or posting time. No one can guarantee virality, but you can raise your odds.
Most people think a Reel goes viral because the creator is famous. The evidence points elsewhere. Instagram shows most Reels to viewers who don't follow the account, so every Reel is tested on strangers first. The ones strangers finish and forward get shown to more strangers. This guide explains that loop, the signals behind it, and how to build Reels that pass each round.
In this guide
- How Instagram decides which Reels spread
- The first three seconds: skip rate and retention
- Sends: why forwarding beats likes
- Originality, eligibility and Trial Reels
- Viral vs average Reel, and a 7-step framework
- Common mistakes, tips and FAQ
How Instagram decides which Reels spread
Instagram says there is no single algorithm. Feed, Stories, Explore and Reels each have their own ranking system. In its ranking explainer, Instagram says Reels works much like Explore: it first sources Reels a viewer might like, mostly from accounts they don't follow, then orders them by how interesting it predicts each will be to that person. The key predictions it lists are whether the viewer will watch to the end, like it, find it entertaining, or open the audio page.
The second piece is the test. Instagram has described eligible content being shown to a small audience first, then to wider groups as it performs. Follower count is not the gate; the first small audience is. Instagram has also admitted that its older ranking favored large accounts and aggregators, and it changed this to give smaller original creators a better chance (see Instagram for Creators).
On 22 January 2025, Mosseri named the three signals that matter most. Social Media Today reported that he told creators to watch average watch time, likes per reach and sends per reach in their Insights. Likes count slightly more for people who already follow you. Sends count slightly more for people who don't.
| Signal | What it measures | Matters most for |
|---|---|---|
| Average watch time | How long people stay with the Reel | Everyone |
| Likes per reach | Likes divided by accounts reached | Followers |
| Sends per reach | DM shares divided by accounts reached | Non-followers |
The "per reach" wording matters. These are rates, not totals, so a Reel seen by 500 people can beat one seen by 50,000 on quality. That is why a niche Reel can still be pushed to a bigger audience.
A popular claim you can't verify. Many guides say sends count three to five times more than likes. ReelDrop's review found no named source for it, and Mosseri's own wording was "slightly", not a multiple. Treat it as folklore. Nobody outside Instagram knows the exact weights.
Speed matters too. Metricool says most Instagram views arrive in the first three days after posting, so the test rounds happen quickly (Metricool, June 2026).
The first three seconds: skip rate and retention
In August 2025 Instagram added a retention chart and a skip rate metric to Reels Insights (Social Media Today). Skip rate is the share of views from people who skipped the Reel within its first three seconds. Instagram says every retention line slopes down, but the flatter the line, the more engaged the audience.
This is the most direct way to see whether your opening works. A high skip rate means the hook failed. A steep drop in the middle means the pacing or payoff failed.
For scale, Metricool's 2026 study of 24,364,803 posts from 375,118 accounts put average Reel watch time at 8.5 seconds, more than double the roughly 4 seconds of a year earlier (Metricool). Viewers are staying longer, so a Reel that holds people past that average has an edge. This is an average across all accounts, not a target.
What a strong opening usually does (editorial guidance, not an Instagram rule):
- Shows something understandable in the first frame, even with the sound off.
- States one specific promise or question right away.
- Skips greetings, logo animations and slow setup.
- Delivers proof or the first payoff early instead of background.
Length and replays. Instagram doesn't give an ideal length. Because watch time is judged per viewer, a short Reel people finish and replay can send a stronger signal than a long one most viewers abandon. One summary of Instagram's published Reels ranking says completion is judged against other Reels of the same length. That could not be confirmed in Instagram's own page, so treat it as a useful idea rather than a fact.
Sends: why forwarding beats likes
A like is public and cheap. A send is a person choosing to spend their own social capital on your Reel. Social Media Today's reading of Mosseri's comments is that sends signal appeal beyond your own audience, because people forward Reels to friends who don't follow you (source). Likes, by contrast, can reflect an in-joke among your existing followers.
Mosseri has also said, as reported by creator-tool blogs, that sends per reach correlate with overall reach more than anything else in his experience. The original Q&A could not be opened, so treat this as reported rather than confirmed.
People are sharing more. Metricool found Reel shares rose 67% year over year, the biggest gain among Reel interaction metrics (Metricool via NetInfluencer).
What makes someone send a Reel. This is an editorial framework, not Instagram data. Most sends fall into three groups:
- Useful: "You need this." A tip, a fix, a warning, a shortcut.
- Relatable: "This is so you." A feeling or situation a specific friend will recognise.
- Surprising or funny: "Watch this." Something that works better when two people see it together.
If you can't name the person who would send your Reel, and to whom, it probably won't be sent.
Worked example (invented numbers). Reel A reaches 12,000 accounts and gets 360 sends. Reel B reaches 60,000 accounts and gets 900 sends. Reel B has more sends in total, but Reel A has the stronger rate: 3.0% sends per reach against 1.5%. Rates are what Mosseri told creators to track.
Comments. Mosseri's top three did not include comments. Instagram's own explainer still counts a viewer's commenting history as part of their activity, so comments are not ignored. They just aren't named as a headline signal.
Originality, eligibility and Trial Reels
Before signals matter, a Reel has to be eligible to be recommended to non-followers.
Originality. On 30 April 2026 Instagram said accounts that mainly re-upload other people's work will not be shown in places where it recommends content. This already applied to Reels from 2024 and now covers photos and carousels too. Instagram says 75% of recommendations in the US now come from original posts. An aggregator account can become eligible again when a significant share of its recent posts are original over a rolling 30-day period, and Account Status in settings shows where you stand. Followers still see your posts either way (Instagram for Creators).
What counts as original: footage you shot, content you designed, and content you materially edited, such as adding context or new information to a meme template or clip. What doesn't: adding only a border, watermark, subtitles or a credit line to someone else's video.
Clean format. Instagram's 2024 post lists eligibility as original content with no visible watermarks that also meets its recommendation guidelines. Export without another app's logo.
Trial Reels. Introduced on 10 December 2024, a Trial Reel is shown to non-followers first and kept off your grid and your followers' Feed and Reels tab. Metrics appear about 24 hours later. You can share it with everyone yourself, or let Instagram do it automatically if it judges the first 72 hours of views to be strong (Instagram for Creators). A follower can still see one through a DM or an audio page. Eligibility rules have changed since launch, so check the option in your app.
Viewer topic controls. In December 2025 Instagram launched Your Algorithm, which shows viewers the topics behind their Reels and lets them adjust them (launched in the US first, per TechCrunch via ReelDrop). Instagram hasn't said how topic choices weigh against watch time and sends. A clear niche is likely easier to match to interested viewers, but that is an inference.
Hashtags. Meta's Reels ranking explainer lists no hashtag signal. Metricool found posts using at least one hashtag got 31.7% fewer views than the platform average, but that is a correlation across all post types, not proof that hashtags hurt.
More competition. Metricool found brands published about 35% more Reels in early 2026 than a year earlier, and a separate Metricool study reported Reels reach down about 35% year over year (Metricool). Millions of views are not getting easier, which is why the hook and the share trigger matter.
Viral vs average Reel, and a 7-step framework
The table below is an editorial framework built from the signals above. It is not measured data.
| Element | Average Reel | High-potential Reel | Signal it targets |
|---|---|---|---|
| Opening | Greeting, logo or slow setup | Clear visual and specific promise in the first seconds | Skip rate |
| Idea | Generic topic or tip | One specific, useful, relatable or surprising idea | Sends per reach |
| Pacing | Padded to reach a length | Tight cuts, as long as the payoff needs | Average watch time |
| Ending | Fades out | Payoff at the end, or a loop that invites a replay | Completion, replays |
| Share trigger | None | A clear person who would want to see it | Sends per reach |
| Originality | Reposted or lightly edited clips | Own footage, own angle, or materially edited | Recommendation eligibility |
| Format | Watermarks, borders, low resolution | Full frame, clean audio, captions | Quality and eligibility |
| Testing | Post and hope | Compare against your own last 10 Reels, test hooks as Trial Reels | Learning |
How to build a Reel that can pass the test rounds
- Pick one idea a specific person would send to a specific friend.
- Write the first line and first frame before you film. If you have access, test two openings as Trial Reels.
- Cut everything before the payoff. Look at your retention chart to find where people leave.
- End on the payoff or loop back to the start so some viewers watch again.
- Export clean: no watermarks, full frame, captions for sound-off viewing.
- Check Insights at about 24 hours: skip rate, retention, likes per reach and sends per reach.
- Compare with your own last 10 Reels, not with a viral outlier. Remake the winner with the same hook structure and a new topic.
Common mistakes
- Judging by total views. Compare rates (skip rate, likes per reach, sends per reach), not raw counts.
- A slow opening. Greetings and logo intros raise skip rate.
- Reposting with a border or watermark. Instagram lists these as immaterial edits, and repeat reposting can cost an account its recommendations.
- Copying a viral Reel exactly. Copy the structure (hook, payoff, share trigger), not the content.
- Hashtag stuffing. Meta lists no hashtag signal for Reels ranking, and Instagram limits captions to five.
- Chasing a universal best time to post. Instagram publishes none. Check your own Insights.
Practical tips
- Keep a simple sheet with one row per Reel: reach, average watch time, skip rate, likes and sends. Add columns for likes per reach and sends per reach.
- Change one thing per test, such as the first line, and keep the rest identical.
- Make a follow-up Reel when one performs well, instead of moving on.
- Check Account Status if non-follower reach looks stuck.
Frequently asked questions
How many views make a Reel "viral"? There is no official number. Compare a Reel with your own baseline and the accounts in your niche.
Do I need a lot of followers to get millions of views? No. Reels are mostly shown to non-followers and tested on a small audience first, so follower count is not the gate. Larger accounts may still have advantages in distribution that Instagram doesn't publish.
Does posting time matter? Metricool says most views come in the first three days, so early performance matters. Instagram gives no best time, so test with your own audience.
Do hashtags help Reels go viral? Instagram's Reels ranking explainer lists no hashtag signal. Use a few relevant ones for context.
Can a reposted Reel go viral? An unchanged repost is likely to be replaced by the original in recommendations, and accounts that mostly repost lose recommendation eligibility. Material edits that add context or information count as original.
How long should a Reel be? Instagram doesn't name an ideal length. A Reel that people finish or replay beats a longer one they abandon.
Is going viral just luck? Partly. Timing, topic and audience mood all play a role, and no one can guarantee a result. The hook, the share trigger and originality are what you control.
Conclusion
Reels with millions of views usually share one pattern: strangers don't skip them, they watch them through, and they send them on. Instagram tests every eligible Reel on a small audience, and each round is passed by watch time, likes per reach and sends per reach. Build for the first three seconds, give people a reason to forward the Reel, make it original, and judge it by your own rates.
Promote your brand with Blogiefy. If you want your brand or expertise in front of a wider audience, Blogiefy publishes guest posts and sponsored articles. Contact us to get started.
Sources
- Instagram for Creators: Rewarding original creators (30 Apr 2026)
- Instagram for Creators: Trial reels (10 Dec 2024)
- Instagram for Creators: Recommendations and originality (2024)
- Instagram: Ranking explained (2023)
- Social Media Today: Instagram shares algorithm insights (22 Jan 2025)
- Social Media Today: Instagram adds retention metrics (24 Aug 2025)
- Metricool: 2026 Instagram Study (16 Jun 2026)
- Metricool: 2026 Social Media Study
- NetInfluencer on Metricool's study
- ReelDrop: How the Instagram Reels algorithm works (Oct 2026)
- TechCrunch: Instagram's Your Algorithm tool (10 Dec 2025)


