If you ask 10 different creators about whether likes matter or not in 2026, you will get 10 different answers. Some will say that likes are not essential; others will say they check their posts every minute for likes. But the reality is something very different and more specific than either side admits, and when you understand it properly, then the way you think about engagement changes.
This blog will explore what likes actually do inside Instagram’s ranking system today, where they still carry weight, and what a like has to look like to move the needle.
Likes still matter, but not the way they used to. Since 2023, Instagram has silently downweighted raw like counts in favor of watch time, shares, and saves. Likes now function more as a supporting thing than a primary one, especially for Reels. The exception is for smaller accounts, where likes still act as a strong early indicator to the algorithm.
In 2019 and 2020, likes were the best signal. A post that got likes in the first hour got shown to more people, full stop. Adam Mosseri and Instagram’s product team walked away from that model knowingly, in stages, for two reasons: like counts were being gamed at scale, and they measured shallow reactions instead of real value.
The current ranking model, based on public statements from Instagram’s engineering team and creator analytics observed across 2024 and 2025, measures engagement roughly like this:
Notice that likes are on the list, not off it. The shift is in the weight, not the presence.
The logic is simple. A like takes half a second and costs nothing. A share is an active activity: the viewer is willing to put your content in front of someone they know. A save is a promise to come back. Both of those signal that the content had actual utility.
Instagram’s ranking team wants content to be engaging that keeps users on the platform and coming back. Shares and saves says better than likes do, so the model learned to favor them.
Likes have not been removed from the algorithm; they have moved into a supporting role. There are three specific places where likes still meaningfully create reach:
For accounts under roughly 10,000 followers, the algorithm has less information to work with. Every signal counts more, and likes are the easiest signal to collect quickly. A Reel that pulls a high like-per-view ratio in the first hour is more likely to be shown in the Explore feed and to non-followers.
There seems to be a floor. Content that gets almost no likes is treated as low-quality and silently suppressed, even if it has a few shares. Likes act as a baseline signal of “this is not spam or broken content” that the algorithm requires to see before it looks at deeper signals.
Brands still look at like counts when vetting influencers, even if the algorithm has shifted. A media kit with a strong average like rate reads as engaged, which influences deal rates. This is external pressure, but it is actual pressure.
Not all likes are the same. Instagram differentiates between engagement that predicts future behavior and engagement that does not. Valuable likes tend to have three characteristics:
A like from a new account with no posts, or a like tapped so fast that the viewer clearly did not watch the video, does not carry the same weight. Instagram’s spam and quality filters have gotten significantly better at telling the difference since 2024.
For accounts trying to create early momentum on a new post, valuable likes for the Instagram algorithm matter because they set the tone for how far the post travels. A post that opens with 50 engaged likes from real profiles in the first hour acts differently in the ranking system than a post that opens with 500 likes from empty accounts.
The tactics that generate the right kind of likes are not surprising, but they are usually ignored:
Likes alone do not move the needle now. They work best when combined with either saves or shares. A post with a high like-to-save ratio is considered high-value evergreen content. A post with a high like-to-share ratio is treated as high-value social content. Both get shown further than a post with likes alone.
If you look at your top-performing posts over the last three months and only weigh likes, you are missing most of the story. Add saves and shares to that view, and the pattern of what actually works often becomes obvious.
One place likes actively hurt reach in 2026: buying them from low-quality sources. Instagram’s spam detection now flags posts with sudden bursts of likes from disposable accounts, and the whole post can be quietly demoted. The account itself can be flagged for repeat behavior. If likes are being sourced, the source matters more than the count.
The quality of engagement is the change that keeps catching creators off guard. The old model rewarded volume; the new model rewards signal.
Ans: No. Whether like counts are visible or hidden to viewers, the algorithm reads the same underlying number.
Ans: Only insofar as it affects brand deals or social proof. For pure reach, ratio matters more than raw count. A 1,000-follower account with 200 likes per post is usually reaching more people per post than a 100,000-follower account with 500 likes.
Ans: Yes. On Reels, likes are weighted lower relative to watch time and shares. On feed posts, likes still carry closer to their historical weight.
Ans: There is no fixed number, but posts that gather engagement steadily for the first two hours perform better than posts that spike in the first ten minutes and then flatline. Steady is a stronger signal than sudden.
Ans: No official confirmation, but the pattern in creator analytics suggests engagement from accounts you interact with regularly carries slightly more weight because those relationships are read as high-signal.
