Cheerful research

What we measured

Cheerful runs real campaigns and measures real creators, so these are our own observations rather than a summary of somebody else's. Every figure ships with its sample size, its date, and the rule we used to exclude rows — including the columns we decided not to publish because the denominator was too thin to mean anything.

What each dataset contains

Influencer rates 2026: how much influencers charge

  • Creator asks by follower tier (1K – 10K, 10K – 50K, 50K – 250K, 250K+), with the 25th, 50th and 75th percentile of each, in US dollars.
  • Brand offers on the same tiers and the same filters, so the opening gap between what a creator asks and what a brand offers can be read tier by tier.
  • Per-deliverable asks for 10 formats, from Instagram Reels and story sets to long-form video and UGC with no posting, unit-normalised before any percentile.
  • How 40 settled deals landed against the brand's highest stated price: 73% at exactly that number, 98% at or below it.
  • How rates are priced (flat fee, per post, CPM, retainer, commission, product only), what paid usage and whitelisting add, and whether engagement rate moves the ask inside a tier.

Source: 905 creator asks, 298 brand offers and 2,356 itemised rate-card lines stated in real brand and creator email threads on Cheerful, measured 16 September 2026.

Instagram engagement rate benchmarks

  • Median and quartile engagement rate for 6 follower bands, from 1K–10K followers to 1M+ followers, each with its own sample size.
  • The same figures for 9 self-declared Instagram categories (athlete, reel creator, personal blog, digital creator and more), with the median follower count of each so category and size are not confused.
  • The spread inside each band — which turns out to be wider than the gap between bands, the finding a single "average engagement rate" hides.
  • The two columns we held back (posting cadence and paid-partnership rate) and the reason each one failed to ship.

Source: 5,924 Instagram accounts from an index of 1,269,277, observed between 2026-08-17 and 2026-09-06.

How we measure

Our own data, not a survey. The rate figures come from prices stated in real negotiations run through Cheerful, recorded with who said each price and what kind of statement it was, so a creator’s ask and a brand’s offer are never averaged into one number. The engagement figures come from Cheerful’s creator index, one observation per account, so an account measured many times cannot weight the median.

A publish gate on every cell. A rate segment is published only with at least 15 quotes from at least 5 different brands, and no single brand above 50% of the rows. A segment that fails is dropped, never rounded or merged into a neighbour — a healthy sample that is really one company’s price list would publish that company’s pricing.

Exclusions are stated, with counts. Instagram returns a placeholder like count for accounts that hide it, which produces a near-zero engagement rate that is an artefact rather than a low score. 864 of 6,788 accounts carried it and were removed before any percentile. Rates outside a one-off USD fee window, retainers and non-USD prices are out of scope for the rate tables.

Ranges, not a single answer. Every table shows the 25th, 50th and 75th percentile and its sample size. We tested a single recommended rate and it was wrong more often than it was right, so the spread is the result, not noise around it.

Known skew. These are Cheerful’s own campaigns and index: weighted toward direct-to-consumer brands and nano to mid-tier creators, mostly in the US. Neither study is a census of the influencer market, and each page says where its sample stops.

Where these numbers are used

The engagement bands are the reference behind the Instagram engagement rate calculator, the band comparison on every creator media kit, and the measured lists in the creator directory. The rate tables are what the negotiation guide quotes. For TikTok, where follower-based rates mislead, the TikTok engagement rate calculator divides by views instead.

Reuse and citation

Both datasets are published under CC BY 4.0: quote, chart or republish any figure, commercially included, with credit to Cheerful and a link to the study page. Cite the measurement date with the number — a benchmark without its dates is not a benchmark.