Every Instagram Reels Algorithm Update in 2026 (Updated Monthly)

Every time the Instagram algorithm changes, three things happen in this order. Reach drops on the long tail of accounts. Creators panic. Ten blog posts get written explaining the change without naming when it shipped or what the actual signal weight is.

This page is the log. Each entry has a date, a primary source link, the change in two sentences, the impact on creators, and a recommended action. The tracker reads top-to-bottom in reverse-chrono so the freshest is first. The last entry is the 2025-Q3 originality update that is still load-bearing for sub-10k accounts. We refresh on the second Tuesday of every month.

If you are reading this because something changed in your reach this week and you want to know whether it is you or them, scroll until you find an entry within 30 days of when your reach changed. If there is no entry, the answer is probably you.

TL;DR#

The four most consequential Instagram Reels changes since September 2025: the September 2025 originality update (penalizes reposted content for sub-10k accounts), the November 2025 signal-weight clarification (Mosseri publicly named watch time, likes per reach, sends per reach as the dominant signals), the February 2026 originality expansion to photos and carousels, and the May 2026 well-being signal (downweights content classified as alarming or pathologizing). Watch time remains the dominant signal across all surfaces.

The signals that matter most in 2026 (cumulative summary)#

Five signals carry the most weight in Reels distribution in 2026, ordered by impact: watch time (high — the time a viewer spends inside a Reel before scrolling), sends per reach (high — DM-shares carry the most weight for reaching new audiences), likes per reach (medium-high — the engagement-quality signal), originality (medium-high — global ranking factor since September 2025), engagement velocity (medium — first-hour engagement decides distribution scope). Niche consistency and well-being filtering are below these as soft caps rather than primary signals.

[TWEETABLE] Watch time, likes per reach, sends per reach. Mosseri named these three in November 2025. Every other "ranking factor" you have read about is a derivative of one of those three.

2026-Q2#

May 2026 — The well-being signal extension#

Source: Meta well-being update notes, May 2026 (anchor link target). Change: Meta extended the well-being classifier originally launched on Facebook to apply across Instagram Reels, including a new flag for content classified as "alarming, sensational, or pathologizing." The flag is a soft distribution cap, not a removal. Impact on creators: Mental-health, attachment-style, parenting, and wellness niches saw the biggest reach effect. Hooks that pathologize a third party (e.g., "5 signs your partner is a narcissist") earn a soft cap; hooks that name a pattern from inside the writer's experience are unaffected. Average reach drop for affected accounts: 18-31% quarter over quarter, recovering when hook framing changes. Recommended action: If you are in any well-being-adjacent niche, run your last 30 hook openers through the three-rule audit in the psychology-niche hooks post. Rewrite anything that pathologizes a stranger.

April 2026 — Trial Reels rolled out to all accounts#

Source: Instagram Creators blog, Trial Reels announcement. Change: Trial Reels, previously limited to a creator beta, became available to all accounts. Trial Reels distribute only to non-followers for 72 hours; after the window, the creator can promote to followers or let it sit in trial. Impact on creators: Trial Reels are a hook A/B testing surface, not a free-reach hack. Real-account 14-day testing showed trial Reels delivered 2.3x non-follower reach but 0.6x engagement compared to normal-published Reels. Best for hook testing and niche-pivot testing; weak for engagement-driven content. Recommended action: Use trial reels when you are testing whether a hook lands; do not use them for community-building posts. Full breakdown in our 14-day test.

March 2026 — Core algorithm update (EEAT and freshness reweighting)#

Source: Indirect — observed across multiple creator-data aggregators (SocialInsider, Dash Social) over the first three weeks of March. Change: Meta did not publicly name this update. Observed effects: posts with first-hand experience signals (named-source quotes, real-account screenshots, original data) gained reach; posts that pattern-matched as AI-generated lost reach. Freshness re-weighting also visible: continuously-updated reference pages (this kind) climbed. Impact on creators: Long-form, AI-drafted Reel scripts started underperforming hand-written ones in the same niche. Creators who shifted to clear first-hand framing recovered within 2-3 weeks. Recommended action: Audit your last 30 scripts for AI tells. The 4-step humanization framework is the protocol.

2026-Q1#

February 2026 — Originality update extended to photos and carousels#

Source: Easter Herald coverage, May 2026 (post-hoc reporting on a January-February rollout). Change: The September 2025 originality update, originally scoped to Reels, was extended to photos and carousels in mid-February. Reposted images, lightly-edited memes, and aggregator accounts saw the largest impact. Impact on creators: Aggregator-style accounts (those that reposted high-engagement content from the niche) lost 40-60% of reach in two weeks. Original-photography accounts and original-carousel accounts gained share. The "sub-10k accounts" cap from the September 2025 update applied here as well. Recommended action: If you have built reach on a repost mix, the format mix has to change. Original carousels that combine your own data, your own writing, and your own visuals earn the most reach in this niche now.

January 2026 — Watch-time signal weighting (clarification)#

Source: Buffer 2026 algorithm guide and Mosseri Q1 statements. Change: Mosseri clarified that watch time is the highest-weighted signal across all Instagram surfaces (Feed, Reels, Explore, Search) in 2026, having been one of three top signals in 2025 (with likes per reach and sends per reach). The clarification was implicit confirmation that creators optimizing for likes alone were optimizing the wrong number. Impact on creators: Creators who shifted CTAs from "like this" to "save this" or "share this with a friend" saw measurable reach lift within 14 days. Like-bait CTAs underperformed sustainably. Recommended action: Audit your CTAs. Replace "double-tap if you agree" with "save this so you remember next time" or "DM this to the friend who needs it." Watch-time and DM-shares compound on each other in 2026.

2025-Q4#

November 2025 — Public signal-weighting clarification#

Source: Mosseri public posts, late November 2025. Change: Mosseri publicly confirmed that across all surfaces, the three highest-weighted signals are watch time, likes per reach (engagement quality), and sends per reach (DM-shares). The clarification ended several years of speculation about which signals dominate. Impact on creators: Created the foundation for the 2026 metric playbook — hook rate (drives watch time), hold rate (drives watch time), completion rate (drives watch time and engagement velocity), and DM-shareability (drives sends per reach). The 3-metric framework is the operational version. Recommended action: Track the three signals. Stop tracking engagement rate as your north star. Engagement rate measures what your retained audience does; the three signals predict whether new audiences arrive.

September 2025 — The originality update (load-bearing)#

Source: Meta originality update notes (anchor link target). Change: Meta started weighing originality globally as a Reels ranking factor and applied the rule hardest to sub-10k-follower accounts. Reposted content (TikTok crossposts with watermarks, aggregator-style reposts, lightly-edited duplicates) earned a soft distribution cap that drops reach 70-90% versus original equivalents. Impact on creators: The single biggest reach event for sub-10k accounts in the last 24 months. A creator we work with saw last-30-Reel average drop from 12k views to 280 views in 11 days. The fallout was visible on accounts within a week and recovery took 21-45 days for accounts that ran the full reset protocol. Recommended action: If your reach collapsed between September 2025 and February 2026 and you are under 10,000 followers, the 7-day reset is the protocol. Cross-posted TikTok content needs the 4-signal scrub before it can recover.

2025-Q3 and earlier#

Older changes are summarized in the pillar's Pillar 4 section. The originality update is the load-bearing one in the long tail; if you are looking for a 2024 change that explains your current reach, the answer is almost always upstream of one of the four 2026 entries above.

What to do when the algorithm changes (the meta-protocol)#

The instinct most creators have when reach drops is to change everything at once. New posting time, new hook style, new niche, new aspect ratio. The result is no signal. You learn nothing.

The meta-protocol is the opposite: change one variable, hold the rest constant, give it 14 days. Run the 7-point diagnostic to find which check is failing. Run the fix on the failing check, only that check. Watch the next 14 days of data.

The reason 14 days: Instagram's classifier needs about 7-10 Reels of new signal to reclassify an account, and another 4-7 days for the distribution adjustment to land in the data. Anything shorter and you are reading noise.

The other rule: do not panic-post. A creator whose reach dropped 80% and started posting twice a day to "make up for it" usually compounds the problem. Frequency caps fire above 2 Reels per day for most accounts, and the over-posting itself becomes the next reach drag.

How this page works#

The tracker updates monthly, on the second Tuesday. Every entry follows the same five-field template: date, source, change, impact, recommended action. New entries land at the top of the relevant quarter; older entries are not edited, only annotated when newer information surfaces.

Each H3 entry has its own anchor link. If you want to share a specific entry rather than the whole page, right-click the heading and copy the section URL. Each section URL is stable; we do not re-slug entries when refactoring.

We expose RSS at /blog/instagram-reels-algorithm-updates-2026/feed.xml. Subscribe if you want every change pushed rather than poll-checked. The feed is also the source the AI engines consume to keep their citations fresh.

What CreatorHouse does for this#

Tracking the algorithm changes is the easy part. The hard part is figuring out whether your reach changed because of the algorithm or because of your content. CreatorHouse benchmarks your reach against competitor accounts in your niche, so when you see a reach drop, you can immediately tell whether it is global (everyone in your niche dropped, the algorithm changed) or local (only you dropped, your content is the variable).

Frequently asked questions#

What was the most recent Instagram algorithm update?#

The May 2026 well-being signal extension. Meta extended the well-being classifier to flag Reels content as "alarming, sensational, or pathologizing." Mental-health, attachment-style, parenting, and wellness niches saw the biggest reach effect. The flag is a soft cap, not a removal; reach recovers when hook framing changes.

When did Instagram start cracking down on reposts?#

September 2025 for Reels, February 2026 for photos and carousels. The September 2025 update is the load-bearing one — it is the change most likely to explain a reach collapse on a sub-10k account that started in late 2025 or early 2026. Recovery is 21-45 days with the right protocol.

How does the Instagram originality update affect small creators?#

Hardest. The originality classifier applied to sub-10k accounts more aggressively than to larger ones. The mechanism: reposted content, watermarked crossposts, and lightly-edited duplicates get a soft distribution cap that cuts reach 70-90%. The fix is original-format content for the next 30-45 days; the classifier reclassifies once it has clean signal.

Did Instagram change the Reels algorithm in 2026?#

Yes, four times. The January 2026 watch-time clarification, the February 2026 originality extension, the March 2026 EEAT-and-freshness reweighting, and the May 2026 well-being signal. Plus the April 2026 trial-reels rollout to all accounts (a feature change rather than a signal change). The full log is the body of this post.

How often does Instagram update the algorithm?#

Continuously, but the visible step-changes happen 4-6 times per year. The smaller adjustments (signal-weight nudges, classifier retrains) happen monthly and rarely get announced. The tracker on this page logs the visible changes; the smaller adjustments show up in the cumulative-signal-summary section as it gets updated.

Where to start#

If your reach changed this month, scroll to the most recent entry. If your reach changed in a specific window, find the entry closest to that window. If you cannot find an entry within 30 days of when your reach changed, the issue is your content; run the 7-point diagnostic.

The next entry will probably ship before this paragraph does. Bookmark this. Come back.

— Salah

Updates#

• 2026-05-26: Initial publication. May 2026 well-being signal added.