Public opinion analysis on social media

For public bodies, institutions and political parties.

What citizens think about a topic and how that conversation changes. We analyse the videos and comments where public opinion forms, with our own AI checked by people and official access to Meta and TikTok.

  • Conversation peak
  • Narrative shift
Illustrative example. It does not correspond to a real topic or study.

Citizen sentiment on social media

The same method works for a crisis that lasts days and for a policy debated for months.

  • Crises

    Sensitive topics and moments when the conversation concentrates and gets distorted.

  • Campaigns and policies

    How a campaign or a policy lands in what people post and comment.

  • Narratives

    Which stories circulate, who keeps them going and how they link together.

Disinformation detection and misinformation monitoring

We do not rule on what is true. We identify the videos that shift the narrative and document how they circulate. Your team decides with the material in front of them.

  • Which pieces shift the narrative

    The videos and posts that change the course of a conversation, with real examples.

  • How it spreads

    How a narrative circulates and gets amplified, and which other topics it connects to.

  • Documented, not summarised

    Each project is documented with its pieces and its path. It is not reduced to a percentage.

Public opinion for political parties and election campaigns

Which topics drive the conversation during a campaign, which pieces shift it and how each proposal is discussed. With public data and the same method for every candidate.

  • Campaign topics

    Which issues gain and lose weight in the conversation, week by week.

  • How proposals land

    How a proposal or a debate is discussed in the videos that circulate.

  • Hoaxes and misinformation

    Which hoaxes and fake news circulate about a candidate and how they spread.

Beyond sentiment analysis

Positive, negative or neutral does not explain why a conversation changes.

  • Emotion, not just polarity

    Which emotion shows up, not only whether the tone is good or bad.

  • What happens inside the video

    The format and narrative of the video, not only the text of the comment.

  • Checked by people

    The models are checked against a hand-annotated set before they are used.

Real-time topic monitoring

Three ways of looking at time. Pick one or combine them.

  1. A picture of the past

    The corpus of a closed period. It explains what happened.

  2. A current picture

    The state of the conversation at the time of the request.

  3. Real-time monitoring

    The same question, in real time, while the topic is still open.

What is delivered: report, dataset and dashboard

The work does not end in a slide deck.

  • Illustrative

    Report

    The reading, with the method and the limits of the data.

  • Illustrative

    Dataset

    The structured data, so another team can keep working.

  • Illustrative

    Dashboard

    A view to check the result without reopening the study.

Public opinion analysis, frequently asked questions

Which sources do you use?

Public data from Instagram and TikTok, with official access to Meta and TikTok. Sources and coverage are documented in the report.

How do you tell disinformation from opinion?

We do not rule on what is true. We identify the pieces that shift the narrative and document how they circulate. The criteria are agreed with your team before we start.

Is the monitoring real-time?

Yes, while the topic is still open. You can also ask for a picture of the past or of the current moment.

How is the data handled?

We work with public data and a documented protocol. If your organisation has its own requirements, we review them before we start.

Is there a published price?

No. The format and the budget depend on the question and the deadline.

Do you work with political parties?

Yes. We analyse the public conversation with the same method we use for a public body. Public data, criteria agreed before we start and documented limits.

How do you track fake news and hoaxes?

We follow the pieces that circulate about a topic and document how they spread and who amplifies them. What counts as a hoax or fake news is agreed with your team before we start.

Tell us the topic and the deadline.