The graph is the part most creators ignore. Instagram’s dashboard will show that a video held its audience for thirty seconds, then shed a third of its viewers in the next ten, but it will not say why. A creator can see the drop; she cannot see what caused it. The assistant Instagram added to its Edits app this week is an attempt to answer that question in plain sentences.
Instagram put the conversational assistant inside Edits, the editing application it launched early last year, on September 30. The tool reads a creator’s own numbers, follower growth, views, video retention, likes and shares, and blends them with what people are writing in comments, what is trending on the platform and what the audience appears to prefer. The point is to replace generic advice with recommendations grounded in one specific account’s data.
Brett Westervelt, the Meta veteran who leads Edits, put the ambition more concretely. The assistant can point to the exact segment where a video starts to shed viewers. “Ask why one reel held attention and another didn’t,” he wrote in a post announcing the tool. “Ask what’s trending on the platform. Ask what to make next.”
The framing matters because of what Edits is up against. The app was introduced in January 2025, days after the United States briefly banned CapCut, the ByteDance-owned editor that had become the default tool for short-form video. Adam Mosseri, the head of Instagram, pitched Edits as a native alternative built for creators, and Westervelt was later named its first head with a brief to keep people cutting and publishing inside Meta’s apps rather than exporting footage to a rival.
That contest never really ended, even after CapCut returned. The two products now share a similar surface, but Instagram is betting that the data no outside editor can see is the moat. CapCut can trim a clip; it cannot tell a creator that her followers, specifically, stop watching when she switches to a talking-head shot at the forty-second mark. Edits can, because it is attached to the same account that posts the finished video.
The assistant is also a bet about how creators will want to work. Editing tools have spent a decade adding features, more tracks, more effects, more precision, on the assumption that the hard part is the cutting. Instagram’s argument is that the hard part is knowing what to cut toward, and that a chat window holding your retention curve can be more useful than another slider.
The tool arrives at a moment when Instagram is leaning harder on video. The company has said repeatedly that its growth runs through creators, and that short-form video is the format pulling the most watch time. A tool that keeps a creator posting, and keeps her posting into Instagram’s own editor, is cheaper than paying her to stay. Edits already carries a feature called trial reels, which lets creators test variations of a video against different audiences before publishing, and the assistant extends that same logic into conversation.
The assistant was built alongside creators rather than handed to them finished, and the company’s language leans on that idea. In the announcement, Instagram described a creator who “never thought to vary the length” of her videos until the assistant surfaced the pattern. That is the product’s pitch in miniature: the numbers are already in the dashboard, but the insight was buried until a conversational tool could pull it out and say it back.
The risk is the one that trails every generative assistant inside a social app. It has to be right often enough to be trusted, and it has to be wrong in a way that does not feel patronizing. A creator told her video lost viewers because it was “too long” will close the tab if the real reason was the audio. Instagram said it would keep building the tool with creators, but it did not say how it would measure whether the advice actually works.
Analysts described the move as a retention play dressed as a product feature. The creator economy is crowded with editing apps and analytics dashboards, and most of them are one tap from being replaced. Edits now holds something those rivals cannot offer: a direct line to the platform’s own numbers. That is what Instagram is selling.
What remains unclear is whether the assistant can move beyond diagnosis. Knowing where a video loses viewers is useful; knowing what to make instead is harder, and the difference between the two is where the product will be judged. Instagram did not say when the assistant would reach every creator, or whether it would eventually push edits directly into the timeline.
For now the tool is a conversation starter in both senses. It talks, and it is also an opening move in a longer argument Instagram is making: that the best editor for its platform is the one that never has to guess what the platform wants.
