What is a good engagement rate for influencers?

The number is four different numbers

Every engagement rate is a fraction. The numerator is some set of interactions, the denominator some measure of audience, and neither is standardized. Rival IQ puts the Instagram brand median at 0.36 percent, Socialinsider reports 0.48 percent for 2025 and 0.45 percent in Q2 2026, and Buffer's study of 27 million posts reports 4.3 percent. Two things drive that gap and neither is printed beside the headline figure: which audience base each publisher divides by, and whether the accounts measured are brands or creators.

Instagram figures: 0.36 percent brand median (Rival IQ), 0.48 percent across 35 million posts from 447,613 pages (Socialinsider), 4.3 percent across 27 million posts from 273,000 accounts (Buffer). Buffer's benchmark page never names its denominator, and its own numbers rule out followers: it reports a 3.7 percent rate and a median reach of 1,073 for accounts with 10,000 to 50,000 followers, and a follower-based 3.7 percent at that size would require more interactions than accounts reached.

The numerator disagrees too

Two tools reading the same account differ before either one picks a base to divide by, and Instagram's own API runs opposite to every vendor: it subtracts. Summing interactions flat also discards weighting the platform itself applies. Instagram's ranking explainer says that in Feed 'the five interactions we look at most closely are how likely you are to spend a few seconds on a post, comment on it, like it, share it, and tap on the profile photo', while 'the most important actions we predict in Explore include likes, saves, and shares'. Instagram says separately that popularity signals, meaning how many and how quickly other people are liking, commenting, sharing and saving a post, 'matter much more in Explore than they do in Feed or in Stories'. A save and a like are not worth the same everywhere, and no formula in circulation reflects that.

The API row quotes the documented definition of total_interactions in Meta's Instagram media insights reference.

Instagram benchmarks by follower size

Rates fall sharply from nano to micro, then flatten. Nano accounts run at more than double the micro average on Sprout Social's numbers, and every band above 10,000 followers averages under 1 percent. The 1.59 percent global average gets reprinted constantly and dated rarely. HypeAuditor publishes it directly, measuring 1.85 percent in 2023 against 1.59 percent in 2024, while Sprout Social reprints the same figure and credits Statista. Either way it is a 2024 number and the trend under it points down. Brand accounts are a separate population, so never score a creator against them or the reverse. Rival IQ puts the all-industry brand median at 0.36 percent and its Influencers industry category at 0.576 percent. That same page also prints 0.43 percent for the overall median, and gives the top 25 percent as both 1.05 percent and 1.02 percent, which is a fair warning about how much precision any benchmark actually carries.

Influencer account averages as reported by Sprout Social, which states the formula as (Total Engagements / Total Followers) x 100. Brand medians come from Rival IQ, which reports medians rather than means. The two populations are not comparable.

TikTok is measured against views, and vendors disagree

TikTok distribution runs through the For You feed, so a video's audience has little to do with the account's follower count, and rates are computed per view. Two benchmarks using that identical framing do not agree. HypeAuditor reports 10.3 percent for nano creators under 10,000 followers. Its report prints no formula beside that figure; HypeAuditor's help documentation defines its TikTok rate as likes plus comments plus shares divided by views across the latest 30 posts. Socialinsider, across 2 million videos from 214,507 profiles between January 2024 and December 2025, reports 4.40 percent for accounts with 1,000 to 5,000 followers and 4.00 percent for 5,000 to 10,000. Three things differ and none is printed next to the headline figure: the numerator, since Socialinsider counts saves, which should push its number higher rather than lower; the sample window, 30 posts against a two-year corpus; and the population, influencer accounts against all tracked profiles. Socialinsider's platform-wide rate by views rose from 3.85 percent in 2024 to 4.20 percent in 2025 and fell to 3.85 percent in Q2 2026, and the publisher notes its 2026-labeled report carries 2025 values. TikTok publishes no rate at all: its Research API returns like_count, comment_count, share_count, favorites_count and view_count, so the distance between HypeAuditor and Socialinsider is a distance between two vendors, not two platforms. A view-based TikTok rate and a follower-based Instagram rate are not two readings of the same quantity, so the two cannot be ranked against each other in either direction.

The decline with audience size is real, then it stops

The standard claim is that engagement rate falls steadily as follower count rises. Three datasets say the fall stops. Sprout Social has macro accounts at 0.86 percent and celebrity accounts at 0.94 percent. Buffer has 3.5 percent for 100,000 to 500,000 followers and 5.0 percent for accounts above 1 million, the second highest rate in its table. Socialinsider has TikTok at 3.75 percent for 50,000 to 100,000 and 3.95 percent for 100,000 to 1 million. The drop is steep from nano to micro and roughly flat above it. A peer-reviewed result points the same way on a different axis. Leung and colleagues, in the Journal of Marketing (2022), measured engagement elasticity, which they define as the percentage change in consumer engagement from a 1 percent increase in influencer marketing spend, and found that it 'increases significantly with larger follower size' (b = .019, t = 5.92, p < .001). That is a finding about what each additional dollar buys rather than about interactions per follower, and it cuts against picking small accounts on rate alone. Their data covers 5,835 sponsored posts on Weibo in October 2018, and the authors note that data spanning longer, 'involving more influencers (especially smaller ones), could help generalize' it. Compare inside a follower band and inside a platform.

Impressions-based rates are being retired at the source

An Instagram rate computed against impressions is computed against a metric the platform stopped serving. Meta's Instagram Platform changelog entry dated January 21, 2025 introduced views and listed impressions among deprecated metrics, scoped 'Applies to v22.0+. Will apply to all versions April 21, 2025.' Requests made after April 21, 2025 for media created on or after July 2, 2024 return an error. Instagram's media insights reference documents more than twenty metrics and none is named engagement rate. It publishes counts: likes, comments, shares, saved, reach, views, total_interactions.

Why a media kit figure is usually the flattering one

Reach is a subset of followers for most accounts and impressions are a superset of reach, so one post produces three different rates, and nobody choosing which to publish has a reason to choose the smallest. Buffer's data makes the arithmetic visible: median reach per post for accounts with 10,000 to 50,000 followers is 1,073. Divide the same interactions by 1,073 rather than by a 30,000 follower count and the rate moves by more than an order of magnitude with no change to the content. That is rarely dishonesty. It is an unlabeled formula. Ask for a dated screenshot of the underlying insights instead of a rate: interactions, reach, views and follower count for the last 10 to 15 posts. Ask for the sponsored posts specifically rather than a blended average, because a sponsored post is what you are buying. Then compute it yourself, the same way, for every account on the list.

What engagement rate cannot tell you is whether anyone bought

Engagement rate measures whether an audience is awake. It does not measure whether that audience acted, and the two have different drivers. Hughes, Swaminathan and Brooks, in the Journal of Marketing (2019), found that 'in awareness-building campaigns, high blogger expertise drives engagement while in trial campaigns, less expert (or novice) bloggers elicit greater engagement.' The same paper explains why content aimed at a purchase engages less than content aimed at awareness: trial campaigns 'have a more overt persuasion intent and also a higher hurdle to generate customer engagement.' The creator who posts the best rate against an awareness brief is not automatically the one who moves units. Use the rate as a screen that filters out inert or purchased audiences, then judge the campaign on the action that was paid for: tracked links, unique codes, landing page sessions, post-campaign lift. That measurement is covered in the guide to measuring influencer marketing ROI.

The number can be bought, and buying it is regulated

The Federal Trade Commission's Trade Regulation Rule on the Use of Consumer Reviews and Testimonials, 16 CFR Part 465, was published August 22, 2024 at 89 FR 68034 and took effect October 21, 2024. Section 465.1(j) defines indicators of social media influence as 'any metrics used by the public to make assessments of an individual's or entity's social media influence, such as followers, friends, connections, subscribers, views, plays, likes, saves, shares, reposts, and comments.' That list is the numerator and the denominator of every formula on this page. Section 465.8 makes it 'an unfair or deceptive act or practice' to 'sell or distribute' fake indicators, and separately to 'purchase or procure' them. Both halves carry the same two conditions: that the person 'knew or should have known' the indicators were fake, and that the indicators 'materially misrepresent their influence or importance for a commercial purpose.' The buyer is covered, not just the seller. Enforcement predates the rule. In October 2019 the Federal Trade Commission settled what it called its first-ever complaint challenging the sale of fake indicators of social media influence, against Devumi, LLC, which sold fake followers, subscribers, views and likes. According to the complaint, Devumi filled more than 58,000 orders for fake Twitter followers, and sold more than 800 fake LinkedIn followers to marketing, advertising and public relations firms among other sectors. The order imposed a $2.5 million judgment against owner German Calas, Jr., the amount the Federal Trade Commission alleged he was paid by Devumi or its parent company, with the remainder suspended on payment of $250,000. Verify audience quality before the contract is signed.

How to read a rate someone hands you

Seven checks, in order.

• Ask the denominator first. A 5 percent rate means nothing until you know whether it is per follower, per reach, per impression or per view.
• Ask the numerator and the window. Likes and comments only, over the last 12 posts, is a different measurement from likes, comments, shares and saves over 30, and both get called engagement rate.
• Benchmark inside platform, follower band and industry. On Instagram, 1 percent is weak for a nano account and above the 0.94 percent average Sprout Social reports for accounts past 1 million followers.
• Use the median, not the mean. Per-post engagement is skewed. One hit drags the average up, which is why Rival IQ reports medians rather than means, and one viral video can hold a small account's rate up until it ages out of the sample window. HypeAuditor's window is the latest 30 posts.
• Read the comments, not the comment count. Purchased and pod activity shows up as generic emoji strings, repeated one-word replies, and the same handles under every post. That is the cheapest check here and the only one that inspects the numerator directly.
• Weight saves and shares over likes where the creator will share that data. Instagram names likes, saves and shares among the most important actions it predicts in Explore.
• Pair the rate with absolute reach. A 10 percent rate on 2,000 views is 200 interactions. The rate qualifies the audience. Reach still decides scale.

Frequently asked questions