Ozor Video Best Practices
A comprehensive guide to creating outstanding AI-generated videos with Ozor.ai. This skill encodes the principles, patterns, and prompt formulas that consistently produce professional results — so you spend fewer credits and get better output on the first try.
What Ozor Is
Ozor is an AI video agent that turns text prompts (or uploaded documents) into fully produced, animated videos. It renders each scene as an animated React component at 30 FPS, supports real-time preview, and exports to MP4 at 720p, 1080p, or 4K. It's designed for product launches, marketing, and explainers — not talking-head or long-form content.
Key capabilities:
- Text-to-video from a single prompt
- Document-to-video from PDFs, PowerPoints, images, Google Docs, and URLs
- Brand Kit extraction from your website (logo, colors, fonts)
- Scene-by-scene editing and regeneration
- Templates for common use cases
- Landscape (16:9) and portrait (9:16) formats
The Prompt Quality Principle
The single biggest factor in Ozor output quality is the prompt. A vague prompt forces the AI to guess — and its guesses will be generic. A specific prompt gives it constraints to work within, which is where creativity happens.
Every effective Ozor prompt answers five questions:
- What — the topic, product, or message
- Who — the target audience
- How long — duration or scene count
- What it looks like — visual style, colors, mood
- What it should do — the goal (launch, explain, sell, update)
Prompt Formulas
These are copy-and-customize templates. Replace [brackets] with specifics.
Product Launch
Create a 45-second product launch video for [product name].
12 scenes:
(1) Hook — bold one-line problem statement
(2) Problem setup — show the pain point visually
(3) Solution reveal — introduce [product name] as the answer
(4) Product in action — show the main UI or experience
(5) Feature 1 — [key feature with one-line benefit]
(6) Feature 2 — [key feature with one-line benefit]
(7) Feature 3 — [key feature with one-line benefit]
(8) How it works — quick 3-step flow
(9) Social proof — "[customer quote]" or key metric
(10) Pricing/offer — "[pricing or launch offer]"
(11) CTA — "[your call to action]"
(12) Logo close — logo + tagline + URL
Style: [minimal / bold / dark / light]. Brand colors: [colors]. Target audience: [who].
Social Short (Vertical)
Create a 15-second vertical video (9:16) for [platform].
5 scenes:
(1) Hook — bold text: "[attention-grabbing statement]"
(2) Problem — quick visual of the pain point
(3) Solution — show the product or key idea in action
(4) Proof — [metric, testimonial, or before/after]
(5) CTA — "[short call to action]" with logo
Style: [energetic / calm / funny]. Use large text and minimal background.
Feature Announcement
Make a 30-second video announcing [feature name] for [product].
10 scenes:
(1) Hook — "[feature name] is here" with bold text
(2) Problem — the pain point this feature solves
(3) Old way — how users handled it before (brief)
(4) New way — introduce the feature visually
(5) How it works — step 1 of the workflow
(6) How it works — step 2 of the workflow
(7) How it works — step 3 of the workflow
(8) Key benefit — the single biggest improvement
(9) Availability — who gets it and when
(10) CTA — "[try it now / learn more]" with link
Keep the tone [professional / casual / excited]. Use [brand color] as the primary accent.
Explainer / How It Works
Create a 60-second step-by-step explainer for [topic or process].
15 scenes:
(1) Hook — "[topic] explained in 60 seconds"
(2) Why it matters — one-line context for the audience
(3) Overview — high-level diagram or flow
(4) Step 1 — [first step with visual]
(5) Step 1 detail — key tip or callout
(6) Step 2 — [second step with visual]
(7) Step 2 detail — key tip or callout
(8) Step 3 — [third step with visual]
(9) Step 3 detail — key tip or callout
(10) Step 4 — [fourth step with visual]
(11) Step 4 detail — key tip or callout
(12) Common mistake — what to avoid
(13) Pro tip — one expert recommendation
(14) Recap — summary of all steps in one view
(15) CTA — "[next action]" with link or resource
Audience: [who this is for]. Tone: [simple and clear / technical / conversational].
Use a [light / dark] background with [color] accents.
Release Notes
Create a release notes video for [product name] version [X.X].
2–3 scenes per feature:
(1) Intro — "What's new in [version]" with logo
(2) Feature 1 title — [feature name] with one-line benefit
(3) Feature 1 demo — show it in action
(4) Feature 2 title — [feature name] with one-line benefit
(5) Feature 2 demo — show it in action
(6) Feature 3 title — [feature name] with one-line benefit
(7) Feature 3 demo — show it in action
(8) Recap — all features in a single summary view
(9) CTA — "Try it now" with link
Style: [clean / playful / technical]. Duration: ~30 seconds.
Investor Update
Create a 45-second investor update video for [company name].
15 scenes:
(1) Title card — "[Company] Q[X] Update"
(2) Headline metric — [e.g., MRR] with large number
(3) Growth chart — visual trend over time
(4) Key milestone 1 — [what happened]
(5) Key milestone 2 — [what happened]
(6) Revenue breakdown — key financial data
(7) Customer highlight — notable win or logo
(8) Product update 1 — [what shipped]
(9) Product update 2 — [what shipped]
(10) Team update — key hires or team growth
(11) Roadmap — Q[X+1] priorities
(12) Roadmap detail — top 2–3 initiatives
(13) Key risk or challenge — [brief note]
(14) The ask — [what you need from investors]
(15) Close — logo + contact info
Tone: confident and data-driven. Style: professional, dark background, clean typography.
Tutorial / Walkthrough
Create a step-by-step tutorial for [task or feature].
2–3 scenes per step:
(1) Intro — "How to [task]" with context
(2) Prerequisites — what you need before starting
(3) Step 1 instruction — what to do
(4) Step 1 visual — annotated screenshot or demo
(5) Step 2 instruction — what to do
(6) Step 2 visual — annotated screenshot or demo
(7) Step 3 instruction — what to do
(8) Step 3 visual — annotated screenshot or demo
(9) Step 4 instruction — what to do
(10) Step 4 visual — annotated screenshot or demo
(11) Result — show the finished outcome
(12) CTA — "Learn more at [link]"
Use "Next:" transitions between scenes. Include annotated screenshots where possible.
Audience: [who]. Tone: [friendly / technical / conversational].
Scene Structure Principles
Think in scenes, not paragraphs. Every scene should have one job:
- Hook — grab attention in the first 2 seconds (especially for social)
- Problem — state the pain point the audience recognizes
- Solution — show your product or concept in action
- Proof — metrics, testimonials, before/after, social proof
- CTA — tell them exactly what to do next
The classic flow that works for almost any video: Intro → Problem → Solution → CTA
For longer videos, extend the middle: Intro → Problem → Solution → Benefits → Proof → CTA
Scene count guidelines by use case (aim for ~1 scene per 3 seconds):
- Social ads: 5–8 scenes (15–30s)
- Product launches: 10–15 scenes (45–60s)
- Explainers: 12–20 scenes (60–90s)
- Investor updates: 12–15 scenes (45–60s)
- Release notes: 2–3 scenes per feature
- Tutorials: 2–3 scenes per step
Format Selection
Choose before you start — changing aspect ratio later regenerates all scenes.
Landscape (16:9) — 1920×1080px
Best for: YouTube, LinkedIn, website embeds, presentations, investor decks, webinars, conference screens.
Portrait (9:16) — 1080×1920px
Best for: Instagram Reels & Stories, TikTok, YouTube Shorts, Pinterest video pins, Snapchat.
Rule of thumb: if it's going on a phone feed, go portrait. If it's going on a screen or webpage, go landscape.
Brand Consistency
Set up a Brand Kit before creating videos. Ozor can auto-extract brand assets from your URL (logo, colors, fonts). Once set, every video inherits them automatically.
Upload assets before writing your prompt — the AI can only use what's already available. Reference uploaded assets by name in your prompt: "use the product screenshot I uploaded."
Editing Workflow
Ozor is conversational — you refine by chatting, not by dragging timelines.
The most effective editing pattern:
- Start broad — get the structure and flow right first
- Target one scene at a time — click the scene thumbnail, then ask for changes
- One change per message — stacked requests lead to mixed results
- Ask for alternatives — "give me 3 headline options for this scene"
- Save what works — save good scenes as templates for reuse
Useful editor commands:
"Edit scene 2 to..." — target a specific scene"Add a new scene after scene 3" — insert at a position"Make the text larger in scene 1" — adjust visual details"Change the color palette to..." — restyle the whole video"Regenerate scene 2" — redo one scene, keep the rest"Make this scene shorter" — trim duration"Save this as a template" — reuse later"Export at 1080p" — trigger export
Common Mistakes to Avoid
- Being too vague — "make a cool video" gives generic results. Always specify duration, audience, style, and scene structure.
- Not targeting a scene — click the scene thumbnail before asking for changes, otherwise the AI may regenerate the whole video.
- Uploading assets after the prompt — upload images and brand assets first, then reference them.
- Requesting too many changes at once — one change per message. Get the text right, then refine animation.
- Skipping the Brand Kit — set it once, benefit forever.
- Wrong aspect ratio — pick 16:9 or 9:16 at the start. Changing later means regenerating everything.
Export Reference
| Quality | Available on | Best for |
|---|
| 720p | Free | Drafts, previews, internal sharing |
| 1080p | Pro ($29/mo) | Social media, landing pages, presentations |
| 4K | Business ($69/mo) | Conference screens, broadcast, premium campaigns |
Exports render server-side in 2–5 minutes depending on length and complexity.
When Helping Users Write Ozor Prompts
When a user asks for help creating a video with Ozor, follow this process:
- Clarify the use case — what type of video? (launch, explainer, social ad, etc.)
- Identify the audience — who will watch this?
- Determine the format — landscape or portrait? What platform?
- Draft a structured prompt using the appropriate formula above
- Include specific visual direction — colors, style, mood, background
- Suggest a scene count based on the use case guidelines
- Remind about Brand Kit and assets — upload before prompting
Always produce a complete, copy-pasteable prompt the user can drop directly into Ozor. Don't give abstract advice — give them the actual prompt text.
Example: From Vague to Great
User says: "I need a video for my app"
You produce:
Create a 45-second launch video for [App Name], a [one-line description].
12 scenes:
(1) Hook — "[pain point the audience feels]"
(2) Problem — show the frustration visually
(3) Solution reveal — introduce [App Name] with logo
(4) Product in action — show the app's main screen with [key feature]
(5) Feature 1 — [benefit 1] with quick demo
(6) Feature 2 — [benefit 2] with quick demo
(7) Feature 3 — [benefit 3] with quick demo
(8) How it works — 3-step setup flow
(9) Social proof — "[customer quote or key metric]"
(10) Pricing — "Free to start" or [pricing info]
(11) CTA — "Try it free at [url]"
(12) Logo close — [App Name] logo + tagline
Style: dark background, clean typography, [brand color] accents.
Audience: [target audience].
Tone: confident and modern.
Then ask the user to fill in the brackets — or help them fill them in based on what you know about their product.