Instagram Sends Per Reach Is Now the Biggest Reels Ranking Signal

Most brand Instagram strategies are still built around one question: will people like this? That question is now the wrong one.

Instagram sends per reach — how often viewers privately DM your Reel to someone else — has become the platform's strongest ranking signal for reaching non-followers. Buffer's analysis of 9.6 million posts, updated May 2026, confirmed it. Head of Instagram Adam Mosseri confirmed it. And yet the majority of brand Reels are still structured to collect likes and bait comments. That gap is where reach is being lost.

Instagram Sends Per Reach: What Just Changed

Instagram has always used multiple signals to rank content. What shifted in 2026 is the weighting — specifically, how heavily the algorithm leans on DM share behavior when deciding whether to push a Reel beyond your existing followers.

SyncStudio AI's breakdown of the current Reels algorithm puts it plainly: sends per reach is the strongest signal for reaching non-followers, and Mosseri explicitly stated DM shares carry the most weight for unconnected reach. Not saves. Not comments. Not watch-through rate. The private, person-to-person forward.

This is a meaningful mechanic shift. Likes are a low-friction, passive reaction. A DM share requires someone to actively think of another person, decide this content is worth their attention, and send it. Instagram reads that as high-intent endorsement — the algorithmic equivalent of a referral.

Why DM Shares Carry More Weight Than Public Engagement

The underlying logic is about signal quality, not signal volume.

When someone likes a post, it could mean they found it mildly amusing, they're being polite to a friend's account, or they double-tapped by accident. It is a noisy signal. When someone DMs a Reel to a specific person, there is no accidental version of that action. It means the content was relevant enough to trigger a social recommendation.

Aurelius Media's breakdown of Mosseri's nine confirmed ranking signals ranks sends per reach as "Very High" influence, with Mosseri's own quote: "Sends per reach is now one of the most influential ranking signals on the platform."

Instagram is essentially using private sharing behavior as a proxy for content quality — the same way Google uses backlinks as a proxy for page authority. One strong private signal outweighs dozens of passive public ones.

What the Data From 9.6 Million Posts Actually Shows

Buffer's May 2026 dataset is the most comprehensive public analysis of this shift. Across 9.6 million Instagram posts, the researchers found that sends per reach correlates more strongly with Reels distribution to non-followers than any other engagement metric they tracked.

The practical implication: two Reels with identical view counts can have radically different distribution trajectories depending on their DM share rate. A Reel with 10,000 views and 200 DM shares will get pushed to non-followers more aggressively than one with 10,000 views, 500 likes, and 5 DM shares.

The sends per reach ratio — not the raw number of shares — is what the algorithm reads. So a smaller account with a high send rate on a modest-view Reel can out-distribute a larger account with flat DM engagement.

The Engagement Hierarchy Has Flipped

Here is how the signal priority stacks in 2026 for non-follower reach:

• Sends per reach — highest weight
• Watch-through rate — strong secondary signal
• Saves — meaningful, especially for evergreen content
• Comments — moderate weight, quality over quantity
• Likes — lowest weight of any engagement type

Two years ago, that list was roughly inverted. Brands that built their content strategy around that old hierarchy are now fighting for reach with the wrong tool.

What's Actually Working on the Ground

The formats generating the highest sends per reach are not the ones dominating most brand content calendars.

Content that gets DM'd tends to fall into a few specific buckets. First: hyper-relatable moments that make someone think of a specific friend. Think niche professional frustrations, inside-joke-adjacent scenarios, or extremely specific life situations. The mental model is not "will everyone relate to this?" — it is "will this make one specific person think of another specific person?"

Second: immediately actionable tips. Not "5 ways to improve your morning routine" — but a single, specific trick someone wants to pass along because it solves a real problem their friend has right now. The DM becomes a utility delivery.

Third: content with a genuine surprise or revelation. Something that genuinely shifts how you see something, not just confirms what you already believed. Surprise creates a "you need to see this" impulse.

CMS Wire's 2026 social media trends analysis also points to serialized Reels — episodic content that builds habit — as a format gaining outsized traction. Serialized content generates repeat DM behavior as audiences tag the same friends across episodes, compounding the sends per reach signal over time.

Brands leaning into this pattern: niche B2B accounts with sharp professional commentary, DTC brands producing specific product-use tutorials that feel like genuine life hacks, and creator-led accounts that build recurring characters or formats with dedicated fan behavior.

The Contrarian Read: Optimizing for DMs Is Not the Same as Begging for Shares

Here is where most takes on this topic go wrong: they tell you to add a CTA asking viewers to "send this to a friend."

That advice misses the mechanic. A CTA-driven share is a different behavioral signal than an organic one. The algorithm is reading for unprompted DM behavior — the kind that happens because the content was so relevant or surprising that sharing felt natural, not because you asked.

Asking for shares can produce a short-term sends bump, but it trains your audience to share performatively rather than genuinely. Worse, it makes the content feel like it's working for the brand instead of working for the viewer.

The right optimization is upstream: change what you make, not what you ask for. If you build content with a specific "send this to that one friend" scenario baked into the hook or the premise, the shares happen without the ask. The CTA should never be the strategy.

HeyOrca's 2026 Instagram updates tracker adds another layer here: Instagram's new "Your Algorithm" controls, now available to all English-speaking users globally, let audiences add and remove topics and reset their recommendation history entirely. This means the distribution environment is becoming more fragmented and personalized. Content that gets DM'd into a friend's world is now one of the few organic ways to break through someone's self-curated feed bubble.

How to Increase Reels Reach by Building for Sends Per Reach

These are the specific moves worth shipping this week.

Audit your last 10 Reels with a single question. For each one, ask: "Would someone DM this to a specific friend without being asked?" If the honest answer is no for 8 out of 10, your content brief needs to change before your posting cadence does.

Reframe your hook construction. Stop opening with brand claims or generic setups. Start with a scenario, frustration, or observation so specific that it immediately triggers a person-recognition response — "this is exactly [friend's name]."

Build at least one "utility forward" piece per week. A single, specific tip — not a listicle — that solves a problem someone would immediately want to pass to a peer. Keep it under 30 seconds. Make the value obvious in the first 3 seconds.

Test serialized formats. Two or three connected Reels with a recurring premise gives your existing audience a reason to DM the same friend repeatedly, building a compounding sends-per-reach pattern over weeks rather than requiring each individual post to earn its own cold DM behavior.

Track sends in your native analytics. Instagram shows send counts in post insights. Start logging sends per reach (sends ÷ reach) across your last 30 Reels and look for the content types that index highest. That is your actual signal map — not your like-to-impression ratio.

Stop A/B testing thumbnails and captions first. The biggest lever is the content premise itself. Thumbnail optimization is a single-digit percentage gain. Getting the DM-share mechanic right is a category-level shift in distribution.

What to Watch in the Next 60 Days

Two signals will confirm whether sends per reach continues to dominate or gets rebalanced.

First, watch whether Instagram introduces any sends per reach visibility in its professional dashboard or creator insights in a more prominent way. Mosseri surfacing the metric publicly is usually a precursor to making it more actionable in-product.

Second, watch how the "Your Algorithm" controls evolve. HeyOrca's reporting shows Instagram is actively building user-facing feed controls — if those controls start weighting private sharing behavior in how the system learns preferences, sends per reach becomes even more foundational. The accounts that build DM-share habits now will have an asymmetric advantage as the system matures.

The brands still running last year's playbook — high-production Reels with like-bait captions — are not going to collapse overnight. But they are quietly losing distribution ground every week to accounts that figured out the new mechanic first.

The Takeaway

Instagram sends per reach is not a trend to monitor — it is already the primary lever for Reels distribution to non-followers, confirmed by Mosseri and corroborated across multiple independent analyses of real post data. The content that earns the most reach in 2026 is content someone wants to DM to one specific person, not content optimized for the most passive reaction from the broadest audience. Audit your last month of Reels against the DM-share question, rebuild your brief around specificity and utility, and start tracking sends per reach as your lead distribution metric. Everything else is secondary.