Instagram's DM share signal now drives 3 to 5 times more reach than likes

Instagram head Adam Mosseri confirmed in January 2025 that the three most important ranking signals are watch time, sends per reach (DM shares), and likes per reach. Six months later, that hierarchy has held, and industry data shows DM shares are weighted 3 to 5 times higher than likes for Reels distribution.

One DM share is now worth roughly 15 likes in Instagram's distribution model. A save is worth around 10. The platform has essentially shifted from measuring popularity to measuring intent.

That is not refinement. It is a rewrite of the rules. And most brands are still optimising for a signal that barely matters.

Why Instagram made DM shares the top signal

DM shares are the most heavily weighted ranking signal for reels in 2026, above likes, comments, or watch time, as Adam Mosseri has confirmed publicly. A DM share signals that a viewer found the content valuable enough to send to a specific person.

You cannot bait a share. Either someone cares enough to hit that paper airplane icon and pick a name from their friends list, or they don't. A DM share has a real social cost.

Instagram's internal data showed that DM shares correlate with purchase intent 4 times stronger than likes. That matters commercially. It also explains why the algorithm pivoted so hard. Shares are the only engagement metric Instagram cannot easily game.

What this means for reach

A reel with 5,000 likes and 0 DM shares will reach a fraction of the audience of a reel with 1,000 likes and 50 DM shares. A post with 50 likes and 50 shares will outperform a post with 500 likes and 2 shares.

The shift shows up across every surface. When someone DMs your Reel to a friend, Instagram treats that as a strong quality endorsement and accelerates distribution, with sends weighted 3 to 5 times higher than likes for reaching new audiences. Shares through DMs from people who don't follow you carry extra weight on Explore.

The algorithm evaluates a post in three windows. Window one is the first 20 minutes where early engagement signals decide if the post gets pushed beyond your followers. Window two is 20 minutes to 6 hours when the algorithm tests the post with a small non-follower audience. Window three is 6 to 48 hours where strong-performing posts get a second push to the Explore feed. Posts that fail window one almost never recover in windows two or three.

If your content does not generate DM shares in that first window, it is capped before it ever reaches non-followers.

The trade brands must make

The question to ask about every piece of content before publishing is: would someone send this to a friend? If the answer is no, the content is unlikely to break out of your existing follower base.

That creates a tension for brand accounts. For brands managing creator content, this update creates a clear brief: creative freedom matters more than brand control. Content that looks like it was made for someone's personal page gets shared privately. Content that looks like an ad mostly gets skipped.

The algorithmic incentive is clear. The brand control reflex is the opposite. That is the central problem social teams are navigating in July 2026.

How professionals should respond

The obvious response is to design content that prompts shares. DM shares happen when viewers think a specific friend needs this right now. Tactical content, named lists, before-and-after reveals, and 'I cannot believe this' moments drive DM shares. Generic educational content gets saves but not shares. Be specific enough that one viewer pictures one friend who needs this.

Creating shareable content such as relatable takes, useful resources, and niche humour is the most effective strategy for growth in 2026. Content that gets shared in DMs receives significantly broader distribution than content that only gets likes.

Structural changes help. Carousels are having a significant moment, with engagement rates up over 30 per cent year on year. The mechanic is simple: carousels keep users swiping, which extends time on post, which Instagram reads as a positive quality signal.

But the format is secondary to the insight that makes someone want to forward it.

What did not change

Watch time remains the top signal overall. Watch time is the number one signal, followed by DM shares (sends per reach), then likes. Comments didn't even crack the top three.

Watch time is the single most important signal. Instagram tracks total seconds watched, percentage of video completed, and whether viewers rewatch. If most viewers watch past the 3-second mark, Instagram interprets that as a working hook and pushes the Reel wider.

The first three seconds still decide whether a Reel gets distribution or dies. Data suggests that up to 50 per cent of viewers drop off within the first three seconds. If you do not capture attention immediately, the algorithm reads that as a signal your content is not worth distributing.

Shares amplify reach. Watch time unlocks the test. Both matter, but they serve different functions in the ranking system.

The commercial consequence

For social media professionals, this is not a tactical shift. It changes the brief. The job is no longer to produce brand-safe content that hits KPIs. The job is to produce content so useful, surprising, or relatable that someone forwards it to a specific person.

That is harder to control. It is also the only route to meaningful non-follower reach on Instagram in July 2026. Accounts with smaller but highly engaged followings are routinely outperforming large accounts with passive audiences. The algorithm reads engagement signals, not follower counts. If your content drives watch time, sends, and likes per reach it will be distributed, regardless of whether you have 500 or 500,000 followers.

Every brand optimising for vanity metrics is now invisible to the algorithm that decides who sees their work. That is the underlying problem. The DM share signal is simply the mechanism that surfaced it.