Insights · Platforms ·

The replies under your post are not the public

A new report on six US midterm races found that up to one in three comments on a candidate's own X account came from inauthentic accounts. If your team reads its replies as a mood check, it is reading the most distorted sample there is.

Every comms team has a habit it rarely questions: post something, then read the replies to see how it landed. A new report suggests that habit is getting worse at telling you anything.

On October 28, the analytics company Cyabra published a study of six competitive US midterm races: the Senate races in Kansas, Maine, and Michigan, and the governor's races in Georgia, Nevada, and Ohio. It looked at about 274,000 posts and comments on X, TikTok, Facebook, and Instagram between August 31 and October 19, 2026.

What it found

  • About one in nine posts about the candidates came from accounts Cyabra classed as inauthentic, at roughly the same rate for Republicans and Democrats.
  • The rate varied by platform: about one in four on TikTok, one in seven on X, one in eleven on Instagram, and under one in twenty on Facebook.
  • Replies were the worst. On the candidates' own X accounts, between one in three and one in five comments came from inauthentic accounts in most races. That is up to three times the overall rate.
  • Coordination showed up in five of the six races, and in four of them both candidates were hit.

A caveat: this is a vendor's report on its own method, published in a press release. Treat the exact ratios as one estimate. The pattern is what matters, and it matches what organizers already feel: the space right under your post is where pressure concentrates.

Why this matters for your team

When a party, union, or NGO posts, the replies feel like the public talking back. They are not. They are a small, self-selected group, and according to this report a large share of it may not be people at all. A team that adjusts its message, apologizes, or launches a counterattack because "the replies are on fire" can end up steering by a crowd that was built to steer it.

Replying makes it worse. Replies and quote posts count as engagement, and engagement buys reach. Answering a hostile thread under your own post shows it to more people, including people who would never have seen it.

Three habits that help

  1. Stop reading mood from your mentions. Look at what your own members and allies are saying in their own channels, and at whether the topic is growing outside your replies. If it isn't, it is noise.
  2. Don't reply to bait. Scrolling past is a choice your team can make on purpose. Most attacks fade faster when nobody from your side gives them oxygen.
  3. Fill the first hour with real people. Ranking systems decide early how far a post travels. Ask your members to share, send it to a friend, and add a sentence in their own words in that first hour. Real people with names and faces are the one thing a reply farm cannot copy.

Where Ostracon fits

Ostracon doesn't hunt bots and doesn't tell you who is real. It helps with the decision that comes before any response: whether to answer at all. Its briefing recommends one of three moves, ignore, mirror, or push your own agenda, with ignore as the default on opponent bait. A person on your team makes the call. If the answer is to push, Ostracon sends a mission to your supporters, and they post in their own words.

Sources

Insights

Read next.

Organizing ·

What organizers can learn from Serbia's student vote

A movement with no leader and a list founded a month before the vote won a majority on the highest turnout in a decade. Five lessons for any organization that coordinates many groups at once.

Read

Regulation ·

Europe's platform rules have teeth now. Your response still can't wait for them.

The EU has fined X and opened cases against Meta and TikTok under the Digital Services Act. Enforcement runs in months and years. A false claim about your organization runs in hours.

Read

Ostracon

Act within the hour. Or know when not to.