Thousands comments can mean “we love this”.

They can also mean “what the hell is this?”

Traditional metrics (likes, views, etc) count both exactly the same - and that’s a problem most social reporting has never really solved.

Algorithms don’t see opinions. They see activity - comments, shares, saves - and treat high activity as a signal worth rewarding. A post that provokes confusion, anger, or mockery can outperform a post that genuinely resonates, at least on paper, because the reaction that drives the most frantic engagement isn’t always the positive one.

Which results in brands optimizing for what the numbers reward without always understanding what produced them.

American Eagle’s 2025 “Great Jeans” campaign featuring Sydney Sweeney generated enormous comment activity - but a significant portion of that activity was controversy and debate.

The brand stood behind the work, but the narrative in the comment section had already travelled well beyond what the team could control.

When they brought Sweeney back for a summer 2026 campaign, the original controversy surfaced immediately in the new post’s comment section. Almost a year on, the virality has died down but the narrative is still very much alive.

The comment section has a specific signature depending on which kind of engagement you’re looking at.

Genuine affinity tends to show up as personal stories, unprompted product questions, people tagging their friends, expressions of loyalty to a brand they admire. Negative engagement looks completely different - mockery, corrections, competitor comparisons, expressions of distrust or confusion.

Both produce identical engagement rates, but only one of them is building something meaningful.

Most teams aren’t reading comments with that distinction in mind. They scan for notably positive or negative reactions and treat anything in the middle as noise - which means the dominant narrative, the recurring language, and the emotional undercurrent, often goes unread entirely.

There’s a version of social reporting that treats a high-performing post of evidence of success, full stop. And there’s a version that asks: what were those people actually saying?

The second version is harder. It requires not just reading, but analyzing the comment section like a researcher - looking for what repeats, what surprises, what the audience is processing in public that the brief didn’t anticipate. But it’s the version that tells you something real about how the brand is landing, rather than just how far the post travelled.

Most reporting tells you how a post performed. The comment section tells you what it meant. Those two things are rarely the same — and the gap between them is where most brands are flying blind.

(Which is why Siftsy exists, wink wink)

Algorithms collapsed the chronological feed a long time ago, yet campaign rollouts are still catching up.

In a recent guest essay for Link in Bio, strategist Matthew Stasoff calls the fix “on-ramps” - chronology-agnostic content that brings audiences into a campaign regardless of where they first encounter it.

The thinking: if someone drops into your campaign on day nine with no prior context, does that piece of content still make sense on its own? Does it make them want to see more?

Every post is a potential on-ramp, and the comments underneath it are often the first thing a new audience reads to decide whether the brand is worth their attention.

What the Siftsy team has been reading, saving, and talking about this week.