On 12 November 2026, Buffer published a study of 2.7 million Instagram Reels posted in 2026. Reels of 90 seconds to three minutes reached 4 to 7% more people than the same account’s own average. Reels of three minutes or longer reached 49.3% fewer. For a UK small business: aim for 90 seconds to three minutes when you want reach.
The study is How Long Should Instagram Reels Be? We Analyzed 2.7 Million Reels, written by Kirsti Lang with Buffer’s data science team. We read it in full on 16 November and every figure and quote below is copied from it. The part most coverage will skip is the method. Buffer did not compare big accounts with small ones. It compared each account with itself, which is the only version of this study that tells a 600 follower business anything useful.
What did Buffer find?
Buffer sorted Reels into six length buckets and measured each bucket against the posting account’s own baseline. Reach, which counts the unique accounts that saw a Reel, moved like this:
Under 15 seconds: 1.5% fewer people than the account’s baseline.
15 to 29 seconds: 0.9% fewer.
30 to 59 seconds: 2.7% more.
60 to 89 seconds: 1.6% more.
90 to 179 seconds: 7.4% more.
180 seconds or longer: 49.3% fewer people, and 45.9% fewer views.
The author’s own line on the band that works: “If reach is your goal, I’d suggest 90 seconds to three minutes as your sweet spot.” On everything shorter: “Below 90 seconds, the differences between lengths were small enough to ignore.”
Long Reels were not a dead loss. Buffer reports that “Reels of 180 seconds or more had an engagement rate 2.65 percentage points above their account’s baseline”, the highest of any bucket. Fewer people saw them, and a larger share of those people liked, commented, saved or shared.
Why does a small account’s own baseline matter?
Because it removes the accounts that would otherwise swamp the data. Buffer says: “We compared each account’s performance to itself, so the data wasn’t skewed by Instagrammers with enormous followings”. An account had to have at least three Reels in a length bucket, and to have used at least two buckets, before it counted. Reach and views were log-transformed to limit the pull of one viral post, and engagement rate was capped at the 99th percentile because the raw data held broken values.
In plain terms, the 7.4% figure is the average lift a given account saw when its own Reels ran 90 seconds to three minutes, compared with its own usual result. Buffer adds that the median account saw about 4%, which is why the study quotes a range of 4 to 7% rather than one number. Buffer also says the finding is an association, not a guarantee. Making a Reel longer does not add 7% on its own.
Should a UK small business make longer Reels?
Yes, up to three minutes, if the Reel has three minutes of substance. The job that feels this first is whoever edits the Reels: the marketing manager or owner at a 10 to 50 person consumer facing firm (a gym, a restaurant, a kitchen fitter, a boutique), or the freelancer who runs social for four of them. The Reel formats that already run long for these businesses, a job walkthrough, a before and after, a dish being made, a tour of the premises, are the ones that fit the band without padding.
The caveat is the sample. These are Reels published through Buffer, a scheduling tool, between January and mid-August 2026. Buffer gives no country split, so the UK is in there but not separated out, and the accounts skew towards businesses and creators who schedule rather than post live. Instagram also changes how it distributes Reels often, so the band could move. Treat the result as the best available evidence, not a rule.
Worked example: a twelve person Hampshire landscaping firm posts two Reels a week to about 1,400 followers. The details here are illustrative.
Its current Reels are 20 to 40 second clips of finished gardens with music. The owner films a two minute walkthrough of one job instead: the brief, the problem with drainage, the fix, the result. Same crew, same phone, one extra hour of editing. On Buffer’s numbers the firm can expect the walkthrough to reach around 4 to 7% more people than its own average Reel, which at this size is a few dozen extra accounts per post, not hundreds. The gain is small. The reason to do it is that two minutes is also long enough to show the work properly, which is what turns a view into an enquiry.
The timing side of the same question is covered in our guide to the best time to post on LinkedIn in 2026, and the reposting rules that now apply to short video on YouTube are in YouTube Shorts cuts reach for reposts.
Is a Reel over three minutes ever right?
Yes, when you want depth from existing followers rather than reach. Buffer’s long bucket lost almost half its reach but had the highest engagement rate, and the author’s reading is blunt: “Somebody who sits through four minutes of your video has already opted in.” A detailed product demo, a client case study or a founder answering real questions can earn that. Expect fewer people to see it and plan for that in the caption and the follow up.
One practical check before you try. Buffer notes that Instagram’s longer upload limit “rolled out gradually, so a few accounts may still see a 90-second or three-minute cap”. Open the Reels composer on the business account and look at the limit it shows you before you plan a long edit.
What to ignore
Ignore any post that turns this into “make every Reel three minutes”. Buffer’s own figure for the band is a single digit lift, and the study says a longer Reel is not a guarantee of more reach. A 25 second clip that lands its point will still beat a two minute Reel that pads to hit a number. Ignore the engagement rate figure as a reason to go long by default too. It describes a smaller, keener audience, which is a different goal from reach.
What to do this week
Open Instagram Insights on the business account, sort Reels by reach, and note the length of the top five and bottom five. Check whether your own pattern matches Buffer’s before changing anything.
Pick one format that naturally runs long, a job walkthrough, a how we make it, a premises tour, and cut the next one to between 90 seconds and three minutes.
Check the length limit the Reels composer shows on your account, and leave anything you want seen by new people under three minutes.
Frequently asked questions
What is the best length for an Instagram Reel in 2026?
Buffer’s November 2026 study of 2.7 million Reels found that Reels of 90 seconds to three minutes reached 4 to 7% more people than the same account’s average. Reels of three minutes or longer reached 49.3% fewer people. For reach, 90 seconds to three minutes is the band to aim for.
Do short Instagram Reels still work?
Yes. In Buffer’s 2026 data, Reels under 90 seconds reached within about 3% of the account’s own baseline in every bucket, and Buffer says those differences were small enough to ignore. A short Reel that makes its point is fine. The clear loss only appears past three minutes.
Does the Reel length finding apply to a small business account?
Buffer compared each account with its own baseline rather than with large accounts, so the result describes what happened to an account when it changed length. For a small UK business account the lift is modest, about 4 to 7% on reach, and Buffer calls it an association rather than a guarantee.
Where Lahat fits
Lahat Creative runs social media and content for UK businesses, from the MINI Challenge Racing Series (156% reach and 160% follower growth) to MotoFest (5,000+ attendees, 2M+ impressions and a 214% rise in visibility). Lahat Creative was named PR Agency of the Year 2025/26 at the Corporate LiveWire Global Awards and Best Integrated PR, Events and Media Agency 2026 at the UK Enterprise Awards. If your Reels are not reaching new people, we will look at the last ten with you and say what to change.