Skip to main content
Feature · UGC · Product placement

AI Product Placement Video Generator
your creator, your product, on camera.

Lock a virtual creator, drop your product photo, type a script — Shortlify renders a face-cam UGC video where the creator literally holds and demos your product. Hand gestures, head motion, expressive lipsync, your studio backdrop. A new variation is a script edit and a re-render away — nobody to book.

Demo: virtual creator holds and demos a skincare serum, full UGC product placement pipeline (first frame Nano Banana → per-scene Veo body motion → Sync v3 lipsync → ffmpeg crossfade).
What you get
  • ✓Virtual creator holds, points to and demos your real product on camera.
  • ✓Identity-locked actor — same face/style across every variation you produce.
  • ✓Studio backdrops adapt to the product (skincare → vanity, tech → desk, fashion → mirror, fitness → gym).
  • ✓Expressive body motion: hand gestures, head tilts, hair flow, micro-expressions.
  • ✓Sync v3 lipsync handles photoreal AND stylized creators (Pixar, anime, illustration).
  • ✓Product texture stays clean — our role-aware multi-ref pipeline never bleeds the product pattern onto the creator's clothing.
  • ✓Every variation reuses the same locked creator and product — re-render, no re-shoot.
How it works
  1. 01
    Step 1

    Lock your creator + product

    Pick a creator from your asset library (or upload one photo). Drop your product image into the Product slot. Optionally add a place (your studio) or an outfit. Tag them with @ in the prompt later if you want to control where they appear.

  2. 02
    Step 2

    Describe the studio + write the script

    Tell Shortlify what the studio looks like in plain text — "cozy beauty studio, AcmeSerum on the table, soft golden hour light through the window". The first frame composes automatically (Nano Banana edit, multi-ref). When you're happy, write the UGC script — or hit Generate and Claude writes one in your voice.

  3. 03
    Step 3

    Render and ship

    Click Generate. Behind the scenes: Claude director splits your audio into 3-6 visual beats (different gestures, expressions, head angles), Nano Banana regenerates a keyframe per beat, Veo / Kling adds body motion, Sync v3 repaints the lipsync, ffmpeg crossfades it together. Output: a 30-60s 9:16 MP4 ready for TikTok / Reels / Shorts.

Comparison
ElementShortlify AI Product PlacementReal UGC creator hire
Time per video10–15 min from idea to MP45–7 days (briefing + recording + edits)
VariationsAs many as your credits allow, identity-lockedA new shoot per batch, a different look each time
Cost per video~$2–3 at scale$50–250 per clip (US-based UGC creator)
Product accuracyClean, role-aware multi-ref — no texture bleedDepends on lighting + product handling
Iteration speedEdit script → re-render in 5 minRe-shoot, re-edit, re-deliver
Brand controlLock creator/voice/studio across campaignsNew creator each time = inconsistent
Prompt ideas that work
“Marc demos AcmeSerum face-cam, calls out 3 reasons it works, ends with "link in bio".”
“Lina holds the new AcmeSneakers, walks viewer through colors, "code SAVE20" CTA.”
“Sara unboxes AcmePhone17, points to the camera bump, "this is wild" reaction.”
“David in his home gym demos AcmeProteinBar mid-workout, "5 minutes after my run" voiceover.”
“@Marc explains the @AcmeMakeup formula at his @MarbellaStudio, warm tones, gentle smile.”
FAQ
How does Shortlify make the creator hold the product accurately?
We pass the product image as a "held_prop" reference to Nano Banana with explicit instruction: "this is held as a 3D object in the creator's hand — do NOT bleed its texture onto clothing". This role-aware multi-ref pipeline (released April 2026) eliminates the texture-bleed bug that plagues naive multi-ref edits where the product pattern would leak onto the creator's shirt.
Can I keep the SAME virtual creator across 100 product videos?
Yes — that's the core feature. Save your creator to the asset library once. Every job that locks them as the actor regenerates them with their face, hairstyle and overall look preserved. You build a recognizable AI creator persona for your brand the same way real UGC studios contract a recurring face.
What about lipsync — does the creator REALLY look like they're saying my script?
Yes. Sync Lipsync v3 is repainting the mouth region of the body-motion video pass. So the hands gesture and the head tilts (Veo / Kling pre-pass), but the actual lip movement is precision-timed to your TTS audio. Sync v3 also handles cartoon and stylized faces — not just photoreal — so a Pixar-style AI creator works too.
Can the creator wear my brand outfit?
Yes — slot the outfit in the Outfit asset slot (or @-tag a saved outfit in the prompt) and Nano Banana will replace the creator's clothing with your brand piece. Useful for fashion/streetwear brands that want their hoodie or sneakers visibly on the creator.
How does this compare to the avatar tools I have already tried?
Rather than claim anything about someone else's roadmap, here is exactly what this generator does, so you can line it up yourself. You bring three images: any photo as the creator, your product shot, and optionally a studio or outfit reference. All three are composed with role-aware constraints, which is what keeps the product's pattern off the creator's clothing. The creator is identity-locked, so the same face carries across every variation you render, and the product stays in hand through the gestures and the lipsync. If a tool you are comparing against does all of that, compare on output quality; generate one with the signup credits and judge the frames.
Will TikTok / Meta penalize this as AI content?
Both platforms ask you to disclose AI-generated or AI-edited content, and both provide a label for it — check their current creator policies, since the wording changes. Neither has published anything that says a disclosed AI video is ranked down for being AI. What is observable from the outside is ordinary: the videos that do well hook fast and get finished and shared. Disclose, then compete on the hook.
Can I use someone else's face as the creator?
No — only your own likeness or a fully fictional AI character. Using a real public person's face violates our terms and most jurisdictions' personality rights. Generate a fictional creator (we have a "describe and create" flow) or upload your own photo as the actor.
Related reading

Your first AI UGC product placement video, rendered in 15 minutes.

Start creating →

800 free credits on signup · No credit card required