huntitHelp center

The AI-discovery fields

Every post read by AI discovery carries nine fields on a video, eight on a carousel, seven on a single image and six on a text post. Two classify the post; the rest extract what's in it.

Huntit's own reads render in your reports' language. The two classification values stay English, and quoted text stays exactly as it was said or shown — see Huntit and your language.

A — in any extracted field is an honest empty — it means there was nothing to pull (a music-only video has no hook or transcript). It never means zero and never means broken.

A video over the 10-minute AI-analysis limit has no fields at all — the AI was never asked to read it. That is a different thing from an empty field; see AI discovery.

The AI READ block on a video post, with its fields labelled.

The two classification fields

Each field gets exactly one label. A post is never tagged with a combination, even when it mixes intents.

  • Content group — the post's dominant viewer intent: why someone watches.
  • Content type — the structural format underneath it (Tutorial, Quick tips, Storytime, Reaction, Tier list and the like): how the post is built.

A type that only exists as video — a duet, a reaction, a time-lapse — is never given to a carousel, an image or a text post.

Both labels are filters in your posts feed, so you can pull every Educational & Practical post across everything you track at once.

The AI assigns them and they can't be edited by hand. If one reads slightly off on a particular post, judge the post from its content and the summary.

The eight content groups

These are the eight macro-intents, and they are the full, fixed names used across Huntit:

  • Educational & Practical — teaches, clarifies, explains, or demonstrates. The viewer learns something or sees how something is done.
  • Analysis, Opinions & News — breakdowns, takes, commentary, reactions to events. The viewer gets interpretation or a stance.
  • Storytelling & Narrative — story-first content that holds attention through plot, emotion, and character. The viewer follows a story.
  • Entertainment, Comedy & Performance — humour, skits, challenges, trends, music, dance. Fun is the product.
  • Personal & Lifestyle — daily life, vlogs, behind-the-scenes, updates. The viewer connects with a person.
  • Interaction & Social — Q&A, polls, replies, duets, community prompts. The viewer participates or is directly addressed.
  • Products, Offers & Recommendations — reviews, hauls, unboxings, launches, promotions. The viewer evaluates, discovers, or is sold to.
  • Aesthetic, Sensory & Mood — visual satisfaction, ASMR, textures, atmospheric scenes. The viewer experiences rather than learns.

The extracted fields

  • Summary — what happens or what is presented, in neutral factual terms.
  • Value proposition — the primary reason someone would watch: the payoff. The summary says what the post does; the value proposition says why anyone would care.
  • Visual hook — what the viewer sees in the first two or three seconds, or on the first slide: the opening image, not the opening words. On a single image it's the whole picture; a text post has none.
  • Filming setup — how the post was shot, from what is visible: the setup (one continuous take, montage, screen recording, split-screen, static graphics), the setting, then who appears. It answers what Content type hides — a Tool / Product Demo can be a screencast or a person talking — so you can tell whether your team could reproduce it. Videos only: a carousel, a single image and a text post carry none, so the row is absent rather than dashed.
  • Hook — the first one or two spoken sentences of a video, the dominant text of a carousel's first slide or of a single image, or a text post's opening line.
  • Outro — the closing line or call-to-action: the last spoken words of a video, the last slide's text, or a text post's closing line. A single image has none.
  • Transcript — the full spoken words plus on-screen text for a video, the slide-by-slide text for a carousel, or all the text on a single image, in reading order. A text post has none — its words are the post.

On a Facebook post that links to a web page, the read also sees the linked page's title and description, when the post carries them, and the page's preview picture, when there is one. When the post has no words of its own, its hook and outro come from the text on that picture.