Should the Brand or the Creator Post an Instagram Collab.

An Instagram collab post works best when the creator publishes it and the brand accepts the invite, not the other way round.

That is the clearest finding in Emplifi’s study of more than half a million Instagram posts from the first half of 2026. Both kinds of collab beat a brand’s ordinary posts. The ones the creator published came out ahead on interactions and on the shares and reposts that carry a post beyond the people who already follow you.

So the practical question is not whether to do collabs. It is who presses publish.

What the study measured

Emplifi’s amplification analysis covered 520,198 posts across 3,791 Instagram profiles between 1 January and 30 June 2026. Profiles were grouped by size, from XS (under 10,000 followers) to XL (more than 1 million).

Collabs are now normal. More than 73% of the brands in the sample published at least one collab post in that half year.

The comparison that matters is each brand against itself: collab posts set beside the same brand’s standard posts.

Owned and accepted: the two kinds of collab

An Instagram collab post sits on more than one profile at once and shows to each participant’s audience. Emplifi splits them by who started it:

• Owned collaboration: the brand publishes the post and invites a creator to co-author it.
• Accepted collaboration: the creator publishes the post and invites the brand.

On the screen the two can look identical. In the numbers they are not.

The result: accepted collabs carry the reach

In Emplifi’s words, both owned and accepted collab posts “outperform a brand’s standard posts, with accepted collaborations delivering the strongest uplift”.

On shares and reposts the pattern repeats. The study says collab posts “generate more shares and reposts across every vertical analysed”, and that accepted collaborations again deliver the strongest results, reaching more than five times the amplification of non-collab posts in ecommerce and retail.

That last figure is against ordinary brand posts, not against owned collabs. It is the headline figure for one sector, not an average you should plan around.

Where the gap is widest

Three cuts stand out.

• Format. Reels show the strongest gains from collaboration, particularly accepted ones. Owned carousel collabs show the smallest lift.
• Size. Smaller brands see the largest relative jump. Accepted collab posts from XS brands earned five times more interactions than their standard posts.
• Scale. XL brands publish more collab content and are more likely to lead the post themselves. Accepted collaborations still deliver the stronger uplift for them too.

The size finding is the useful one for most of our readers. A brand with a small following has the most to gain from borrowing a creator’s audience, and the study suggests the creator’s own post is how it gets borrowed.

What the data does not tell you

This is vendor research on vendor client data, so read it with its limits in view.

• It covers Emplifi’s clients, not Instagram as a whole.
• The published findings do not explain how interactions or amplification are calculated, or whether results are per post, per follower or a median.
• Emplifi says sample sizes vary by metric and include only profiles that met the conditions for each comparison.
• It does not say, in what it publishes, whether the creators were paid, gifted or simply partners.
• It is careful on cause. On user-generated content it says the data “does not prove that collab posts cause more organic UGC”. The same caution belongs on every result here: brands that collaborate tend to be the more established ones already.

None of that cancels the direction of the finding. It does mean “five times” belongs in a hypothesis, not a forecast.

Why the creator’s post travels further

The study offers only a brief reason: accepted collabs can feel more native to the creator’s established audience. What follows is our reading of why.

A post that starts on the creator’s profile arrives in their followers’ feeds as the creator’s own work, in the voice and format those followers chose to follow. A post that starts on the brand’s profile, with a creator added as co-author, still reads as the brand talking.

Instagram’s own guidance on Reels points the same way. It says that when the people you invite accept, “the number of people who can see your content will increase”. The audience grows through the collaborator. It makes sense that the post grows further when the collaborator is the one the audience came for.

How to set up an accepted collab

The mechanics are simple. The order of operations is where brands get it wrong.

1. Agree the idea with the creator first, then let them shoot and edit it in their own style.
2. The creator publishes from their own account and adds the brand as a collaborator.
3. Someone on the brand side accepts the invite, on the right account, the same day.
4. The post now sits on both profiles.

Instagram’s Reels guidance says a reel can carry up to five collaborators. That is a ceiling, not a target. Every extra name dilutes who the post belongs to.

The weak point is step three. Until the brand accepts, the brand is not on the post at all, and it is not yet the accepted collab the study measured. Somebody has to own the brand account’s notifications on launch day.

What goes in the brief

Most collab briefs hand over the content: the product, the message, the mandatory lines. Then the brand posts it.

The data argues for handing over the post itself. A brief for an accepted collab should say:

• Who publishes: the creator, from their account.
• Who accepts, and how fast: a named person on the brand side, inside the agreed hours.
• What the brand needs in the post, kept short: the product shown clearly, any claims checked, any required disclosure.
• What the creator controls: the hook, the edit, the caption voice, the posting time.
• The format: Reels first, since that is where the study found the largest gains.
• What gets reported back: shares and reposts, not only likes and comments.

The fewer brand fingerprints on the edit, the more the post reads like the creator’s work. That is the thing the numbers appear to reward.

If your team does not have the process to hand over control and still protect the brand, that is a workflow problem before it is a creative one. Our Facebook and Instagram work is built around that kind of system.

When the brand should still publish

Owned collabs are not wrong. They outperformed standard brand posts in the study, just by less.

We would put the choice in operating terms:

• Accepted, when the goal is new people seeing you.
• Owned, when the goal is control: a launch with exact claims, a carousel that needs product detail, a regulated message.

Emplifi notes the sectors already split this way. Electronics and software brands lean on accepted collabs, fashion brands publish more owned ones, and nearly 60% of travel and hospitality profiles use a mix of both.

How to measure it

Emplifi’s advice here is direct: “Measure shares and reposts, not just likes and comments.”

That fits how collabs work. A like tells you the existing audience approved. A share tells you the post left it.

Imagine a skincare brand running two collabs a month for a quarter, one accepted and one owned, both as Reels. Track for each:

• Shares and reposts, against the brand’s own standard Reels in the same months.
• Profile visits and follows from the post.
• Saves, as a sign of intent rather than a passing tap.

Compare the brand against itself, as the study did. Benchmarks from other brands’ collabs tell you little about yours.

How NBK thinks about collab posts

Collabs are often treated as a creative idea. We treat them as an operating decision: who publishes, who approves, who accepts, and how fast.

Emplifi’s conclusion is that collaborations “work best when brands treat them as part of a broader content, creator, and community strategy”. We agree, and the study gives that a concrete shape. The brand’s job is to choose the right creator, agree the idea and the non-negotiables, then get out of the way of the post.

Choosing the creator is the part numbers rarely settle. Follower counts and engagement rates say little about whether an audience will act on a creator’s recommendation, which is why we look past what a creator media kit tells you.

Next step

If your collabs are going out from the brand account and stalling at the people who already follow you, try the next one the other way round. Let the creator publish, accept the same day, and judge it on shares.

If your team is posting regularly but still feels stuck, NBK can help find the constraint in the system.