Creative and hooks

UGC Ad research: sort UGC Videos and Ads by format and hook

By . Published 1 November 2026. Updated .

A UGC Ad is an Ad that looks like a regular person made it: someone on camera or holding the product, shot on a phone, talking in the first person. Brands pay UGC creators to make them, and most briefs for those creators are written from taste and a few saved examples.

This guide shows how to research UGC Ads by two things you can count: the format of the Video and the type of hook it opens with. It uses a real Research of 46 beauty UGC Videos on TikTok and two UGC-style Meta Ads, all analysed on our own account, and ends with a brief that cites its sources.

Why research UGC by format and hook

Two UGC Videos for the same product can perform very differently. One is a 15 second demo with music. The other is a three minute get ready with me where the creator talks the whole time. Calling both of them UGC does not tell a creator which one to make.

Format and hook tell them apart. Format is the shape of the Video: tutorial, unboxing, talking head, product demo, skit or voiceover over b-roll. Hook is how the first seconds work: a demonstration, a question, a curiosity gap, a problem or a direct address to the viewer. Once every Video in a sample is tagged with both, you can compare groups instead of single Videos.

Organic UGC Videos and paid UGC Ads answer different questions, so research both. Organic Videos show which formats and hooks an audience watches by choice. Ads in the Meta Ad Library show which of them advertisers keep paying for, measured by how long each Ad has run.

What to tag on each UGC Video or Ad

Video analysis and Ad analysis return these fields for every Video or Ad. They are the columns of a UGC research sheet.

FieldWhat it holdsWhat it tells a creator
FormatTUTORIAL, UNBOXING, TALKING_HEAD, PRODUCT_DEMO, SKIT, VOICEOVER_BROLL, VLOG and othersThe overall shape of the Video to shoot.
Hook type and opening lineDEMONSTRATION, CURIOSITY_GAP, QUESTION, PROBLEM, DIRECT_ADDRESS, plus the first line quoted word for wordHow to open, with a real line to adapt instead of a made-up one.
Visual hookWhat is on screen in the first secondsThe first shot to frame.
Face visibilityCREATOR_ON_CAMERA or HANDS_ONLYWhether the creator needs to be on camera or only their hands and the product.
Speaking style and toneCONVERSATIONAL, NO_SPEECH and others, with the main emotionHow to talk: chatty, calm, excited, or not at all.
On-screen textEvery text overlay in orderWhich captions to burn in, and when.
Call to actionVISIT_SITE, COMMENT, FOLLOW, WATCH_MORE, SAVE or NONE, with the lineHow to close.
Sponsorship signals#ad, partner hashtags, paid partnership labels, discount codesWhether the Video was already a paid deal for a brand.
LengthDuration, grouped in Research lift tables (0-15s up to 180s+)How long the cut should be.
Days running (Ads)Start and stop dates from the Meta Ad LibraryWhether an advertiser kept paying for this UGC Ad.

How to research UGC Ads with your Assistant

Each step is a request you can type into Claude, ChatGPT or another Assistant connected to ShortsIntel. Ask for an estimate first on any paid step and it quotes the price without charging.

  1. Research UGC Videos in your category

    Start with one Research on the UGC style you want, not just the product. It finds Videos across your Platforms, analyses up to 50 of them and returns lift tables by format, hook type and length. It costs 1,036 credits, and reading it again later is free.

    Research beauty UGC ad-style creator videos for makeup brands on TikTok. Show me the lift tables by format, hook type and length.
  2. Read lift by format and by hook

    Lift is a group's median views divided by the median of the whole Research. Look for groups with real counts behind them. A group marked low confidence, usually one or two Videos, is something to test, not something to brief.

    Which formats and hook types beat the median, and how many Videos is each one based on?
  3. Pull the opening lines that worked

    Ask for the opening line and visual hook of the top Videos in the best groups, with a template for each. These become the example hooks in your brief.

    For the curiosity gap and direct address Videos, list the opening line, the visual hook and a reusable template.
  4. Find UGC-style Ads that ran longest

    Search the Meta Ad Library for your category, sorted by days running, and keep the Ads written in the first person or with a creator on camera. Meta Ad search costs 1 credit per Ad delivered.

    Search US Meta Ads for fitness apps, longest running first. Which ones are written in the first person, like a customer talking?
  5. Analyse the UGC Ads

    Ad analysis tags a Meta Ad with the same fields as a Video: format, hook, face visibility, on-screen text, call to action and sponsorship signals. A single Meta Ad costs 3 credits, plus 20 credits for the analysis.

    Analyse the three longest running first-person Ads. What format and hook does each use, and is the creator on camera?
  6. Write the creator brief

    Ask for one brief per creator that names the format, the hook type, two real opening lines to adapt, the length and the call to action, and links the Videos and Ads each choice came from.

    Write a UGC creator brief from this Research and these Ads. Cite the source Video or Ad for every choice.

Real output: a UGC Research and two UGC Ad analyses

All three samples come from our own account on the day this guide was written, trimmed to fit. The Research analysed 46 TikTok Videos for the intent below; their median was 179,061 views. The two Ads come from Meta Ad Library searches for fitness apps and meal kits in the US.

TikTok, via REST API

POST /v1/research with intent "Beauty UGC ad-style creator videos for makeup brands", platforms tiktok, max_videos 50

{
  "intent": "Beauty UGC ad-style creator videos for makeup brands",
  "coverage": { "tiktok": "complete" },
  "lift_tables": {
    "set": { "video_count": 46, "median_views": 179061 },
    "visual_format": [
      { "value": "TUTORIAL", "views_lift": 1.36, "video_count": 19, "low_confidence": false },
      { "value": "TALKING_HEAD", "views_lift": 1.18, "video_count": 7, "low_confidence": false },
      { "value": "UNBOXING", "views_lift": 0.96, "video_count": 11, "low_confidence": false },
      { "value": "PRODUCT_DEMO", "views_lift": 0.11, "video_count": 6, "low_confidence": false },
      ...
    ],
    "hook_type": [
      { "value": "DIRECT_ADDRESS", "views_lift": 2.64, "video_count": 4, "low_confidence": false },
      { "value": "QUESTION", "views_lift": 2.59, "video_count": 2, "low_confidence": true },
      { "value": "CURIOSITY_GAP", "views_lift": 2, "video_count": 9, "low_confidence": false },
      { "value": "PROBLEM", "views_lift": 0.92, "video_count": 4, "low_confidence": false },
      { "value": "DEMONSTRATION", "views_lift": 0.69, "video_count": 23, "low_confidence": false },
      ...
    ],
    "length": [
      { "value": "180s+", "views_lift": 2.37, "video_count": 13, "low_confidence": false },
      { "value": "60-180s", "views_lift": 1.63, "video_count": 15, "low_confidence": false },
      { "value": "30-60s", "views_lift": 0.84, "video_count": 8, "low_confidence": false },
      { "value": "15-30s", "views_lift": 0.28, "video_count": 8, "low_confidence": false },
      ...
    ]
  },
  "digest": {
    "hooks": [
      {
        "text": "This is... This is ridiculous. Full face of new makeup launches.",
        "template": "This is... This is ridiculous. Full face of new [product].",
        "hook_type": "CURIOSITY_GAP",
        "why_it_works": "The creator's disbelief builds anticipation for the sheer volume or quality of the products being tested."
      },
      {
        "text": "Deinfluencing K-beauty makeup products.",
        "template": "Deinfluencing [product] makeup products.",
        "hook_type": "PROBLEM",
        "why_it_works": "It uses a trendy buzzword to promise honest, negative reviews, which builds immediate trust."
      },
      ...
    ],
    "gaps": [
      {
        "ask": "Viewers want mystery boxes to offer shade selections based on skin tone so they don't receive unusable products.",
        "angle": "A review of a mystery box that evaluates whether the included shades are universally flattering or a waste of money for certain skin tones."
      },
      ...
    ]
  }
}
Captured 1 November 2026. Tools used: POST /v1/research, GET /v1/research/{id}/digest. Output trimmed; live results change.

Meta Ad Library, via REST API

GET /v1/ads/meta/1198358815416386?analyze=true (a fitness app Ad from a US Meta Ad search for fitness apps)

{
  "ad": {
    "advertiser": { "name": "Train Fitness" },
    "media_type": "video",
    "first_shown": "2025-05-13T07:00:00.000Z",
    "last_shown": "2026-07-10T07:00:00.000Z",
    "days_running": 423,
    "creative": {
      "body": "I was so skeptical but it actually detects keeps track of my exercises accurately!",
      "cta": "Install now"
    },
    "intelligence": {
      "hook": {
        "type": "CURIOSITY_GAP",
        "opening_line": "Cool gym hack. If you have an Apple Watch, you need this app.",
        "visual_hook": "man working out in the gym doing lat pulldowns"
      },
      "visual": {
        "format": "PRODUCT_DEMO",
        "face_visibility": "HANDS_ONLY",
        "on_screen_text": [
          "If you have an apple watch you need this app",
          "It automatically detects what exercise you're doing and counts how many reps you've completed",
          ...
          "It's free to download Go give it a try!"
        ]
      },
      "tone": { "speaking_style": "CONVERSATIONAL", "primary_emotion": "Joy" },
      "persuasion": {
        "sponsored": false,
        "call_to_action": { "type": "VISIT_SITE" }
      }
    }
  }
}
Captured 1 November 2026. Tools used: GET /v1/ads/meta/{library_id}. Output trimmed; live results change.

Meta Ad Library, via REST API

GET /v1/ads/meta/4309638865973296?analyze=true (a grocery delivery Ad from a US Meta Ad search for meal kits)

{
  "ad": {
    "advertiser": { "name": "Hungryroot" },
    "media_type": "video",
    "first_shown": "2026-04-28T07:00:00.000Z",
    "last_shown": null,
    "is_active": true,
    "days_running": 187,
    "variant_count": 12,
    "creative": {
      "body": "This is what happens when I decide I’m done thinking about groceries. Meals, snacks, breakfast, lunch ...",
      "cta": "See details"
    },
    "intelligence": {
      "hook": {
        "type": "CURIOSITY_GAP",
        "opening_line": "POV: This is what happens when you say yes to outsourcing",
        "visual_hook": "man aggressively dropping grocery packages on a counter"
      },
      "visual": {
        "format": "UNBOXING",
        "face_visibility": "CREATOR_ON_CAMERA",
        "on_screen_text": [
          "POV: This is what happens when you say yes to outsourcing",
          "This is only half of it. wait until you see what this actually costs",
          ...
        ]
      },
      "tone": { "speaking_style": "CONVERSATIONAL", "primary_emotion": "Joy" },
      "persuasion": {
        "sponsored": true,
        "sponsorship_signals": ["#HungryrootPartner in caption", "#ad in caption", ...],
        "call_to_action": { "type": "VISIT_SITE" }
      }
    }
  }
}
Captured 1 November 2026. Tools used: GET /v1/ads/meta/{library_id}. Output trimmed; live results change.

What the sample says about UGC formats and hooks

In the beauty Research, the most common hook was a plain demonstration, used by 23 of the 46 Videos, and it came in at 0.69 times the median. Curiosity gap hooks had 9 Videos at 2 times the median, and direct address hooks had 4 Videos at 2.64 times. A creator who opens on the product being applied with no setup is using the weakest common hook in this sample.

Length made a large difference. The 13 Videos over three minutes came in at 2.37 times the median, while the 8 Videos between 15 and 30 seconds came in at 0.28 times. Tutorials, the most common format with 19 Videos, beat the median, and the 6 product demos came in at 0.11 times. This is organic TikTok, so a cut for paid placements may still need to be shorter, but the long version is worth asking the creator for.

37 of the 46 Videos had the creator on camera, and 8 carried sponsorship signals such as a #ad or a partner hashtag. Most were not paid, so they show what the audience watches by choice.

Both Ads open on a curiosity gap, one on camera and one hands only. The fitness app Ad ran for 423 days, and the same first-person line ran on two more Ads from that Advertiser in the same search, at 317 and 285 days. The grocery Ad is a creator partnership, labelled #ad, still running after 187 days with 12 variants listed in the library. Advertisers seldom keep paying for an Ad that does not sell, so both are hooks worth adapting.

A UGC creator brief built from the sample

Every line of the brief points at the evidence behind it, so the creator and the team can check it. This one is written from the beauty Research and the two Ads above.

Brief fieldWhat to writeEvidence
FormatTutorial or talking head, creator on cameraTutorial 1.36 times the median across 19 Videos; talking head 1.18 times across 7; 37 of 46 Videos on camera
HookCuriosity gap or direct address in the first line, not a silent demoCuriosity gap 2 times across 9 Videos; direct address 2.64 times across 4; demonstration 0.69 times across 23
Example opening linesAdapt, do not copy: "This is... This is ridiculous. Full face of new [product]." and "POV: This is what happens when you [decision]"Research Digest hooks; the grocery Ad's opening line
LengthA long cut over 60 seconds, plus a short cut for paid placements180s+ at 2.37 times across 13 Videos; 15-30s at 0.28 times across 8
AngleAn honest review that names what does not work, or a shade check for different skin tonesDeinfluencing hook in the Digest; the mystery box shade gap from viewer comments
Call to actionAsk viewers to comment or visit the site; disclose the partnership with #adAfter no call to action at all (27 Videos), COMMENT (7) and VISIT_SITE (6) were the most common; the grocery Ad's #ad label

Data limits

Data limits for UGC Ad research

  • Research ranks Videos by how far they beat their Creator's usual reach, so a Research sample leans toward Videos that did well. Use lift to compare groups inside it, not as an average for the whole Platform.
  • Meta does not disclose spend or impressions for commercial Ads, so days running and variants added stand in for performance. Variants added can only be counted once an Ad has been seen more than once, which an Advertiser Tracker does.
  • The TikTok Ad Library covers Ads shown in the EU/EEA, the UK, Switzerland and Turkey only.
  • YouTube Creator data covers Shorts only. Video analysis takes YouTube Videos up to 10 minutes long.
  • Hashtags are TikTok and Instagram only. Sounds are TikTok only, and score_video is TikTok only; get_creator scores Videos under 30 days old on every Platform.
  • Instagram and Facebook Keyword search are best effort, and Facebook Video search is opt-in.

What it costs

Each request that does work is a Run, paid from a prepaid Balance of Credits at 1,000 credits per US dollar, with no subscription. The research on this page cost one Research at 1,036 credits and two Ad analyses at 3 credits plus 20 credits each. A Meta Ad search costs 1 credit per Ad delivered, and a single Video analysis 23 credits. New accounts get 1,000 credits free, valid for 90 days. A Run that fails costs nothing, and a partial one is charged only for what finished. Every price is on the pricing page.

For more on how to read Videos and Ads together, see the creative strategy hub and the competitor ad research guide. Every guide is listed on the guides page.

Connect your Assistant

The same server works in every Assistant below. None of them needs an API key: you sign in to ShortsIntel and choose which organization pays.

Claude.ai

Web and desktop. No API key: Claude signs in to ShortsIntel for you.

  1. Open Claude's add connector page with ShortsIntel filled in, and choose Add. Or go to Settings, Connectors, Add custom connector, and paste the endpoint below.
    https://www.shortsintel.com/api/mcp
  2. Sign in to ShortsIntel, approve the connection and choose the organization whose Balance pays for Runs.
  3. In a chat, turn the ShortsIntel connector on and ask your question.

ChatGPT

Web and desktop, in Developer mode. No API key: ChatGPT signs in over OAuth.

  1. Open ChatGPT's connector settings and switch Developer mode on.
  2. Create a connector named ShortsIntel, paste the endpoint and choose OAuth.
    https://www.shortsintel.com/api/mcp
  3. Sign in to ShortsIntel and choose your organization.
  4. In a new chat, choose the + next to the message box, then Developer mode, and turn ShortsIntel on.

Claude Code

One command, then sign in from Claude Code.

  1. Add the server in your terminal.
    claude mcp add --transport http --scope user shortsintel https://www.shortsintel.com/api/mcp
  2. Run /mcp inside Claude Code, choose shortsintel and sign in.

Codex

One command; Codex signs in on first use.

  1. Add the server in your terminal.
    codex mcp add shortsintel --url https://www.shortsintel.com/api/mcp
  2. If no browser window opens to sign in, start it yourself.
    codex mcp login shortsintel

Cursor and VS Code

One-click install links; each asks you to confirm.

  1. In Cursor, open the Cursor install link.
  2. In VS Code, open the VS Code install link.
  3. Sign in to ShortsIntel when the editor asks.

Questions

What is UGC ad research?
UGC ad research is collecting UGC Videos and UGC-style Ads in your category, tagging each one by format and hook, and comparing which groups do better. The result is a brief for UGC creators based on evidence instead of a few saved examples.
Where can I find UGC ads examples?
The Meta Ad Library lists Ads running on Facebook and Instagram, including UGC-style Ads written in the first person or made with creators. The TikTok Ad Library lists Ads shown in the EU/EEA, the UK, Switzerland and Turkey. Organic UGC Videos on TikTok and Instagram show the same formats without ad spend behind them.
How do I tell if a UGC ad is working?
Meta does not publish spend for commercial Ads, so look at how long the Ad has run and whether the advertiser keeps adding variants. An Ad that runs for months is usually one that pays back. For organic UGC, compare a Video's views with the median of a sample, or with its Creator's usual reach.
What should a UGC creator brief include?
The format, the hook type with two or three real opening lines to adapt, whether the creator is on camera, the length, the angle and the call to action. Link the Videos and Ads behind each choice so the creator can watch them.
Which hooks work best for UGC ads?
It depends on the category, which is why you measure it. In our beauty Research of 46 TikTok Videos, curiosity gap and direct address hooks beat the median by 2 to 2.6 times, and plain demonstration hooks came in under it. Run the same Research for your own category before you brief.
How fresh is the data?
Every Research and Ad lookup is dated with the moment its data was read, and you can run a new one at any time. The samples on this page were captured on 2026-11-01.

Try it in your Assistant

Connect ShortsIntel to the Assistant you already use and ask the question this guide answers, or try the calls in the Playground first.