The Official Vidrush Prompting Guide: A Professional Workflow

Introduction: The Flywheel Philosophy

[!WARNING] Fair warning: This guide isn’t for perfectionist and its blah blah for the beginning 😀

If you’re looking to spend weeks crafting a single, flawless video masterpiece, you’re in the wrong place. This is a complete workflow for building profitable, scalable YouTube channels. We’re going to treat Vidrush’s AI exactly as it should be treated: as your personal, tireless production team that’s ready to execute your strategy 24/7.

Here’s the core principle everything revolves around: The Flywheel Effect.

Let me share a reality check about YouTube, 80% of your revenue will come from just 20% of your videos. That’s not a bug, it’s a feature. And once you accept this, the winning strategy becomes crystal clear. Stop trying to make every single video a viral masterpiece (By spending time in the Editor). Instead, increase the number of smart bets you can place.

As Noah Morris puts it: “The trick really is to test fast, fail fast, and see what works.”

How the System Works

You’ll use Vidrush to create a high volume of strategically-targeted, “good enough” videos your 80%. These videos are designed to break even or turn a small profit. Think of them as your data points, your shots on goal.

Here’s where it gets interesting: This volume inevitably surfaces your 20%, those outlier videos that catch fire and go viral. The profits from these winners get reinvested to fund even more content, which produces more winners, which funds more content… You see where this is going. It’s an exponential growth engine that compounds on itself.

Vidrush was built specifically to power this flywheel. It’s not trying to be an pro filmmaker, it’s your idea factory. Until AI can outperform a full human production team on quality, we’re going to win by competing on speed and execution.

The Mindset Shift

This system demands a fundamental change in how you think about content creation. You’re not a director throwing out vague creative notes and hoping for the best. You’re a strategist delivering precise, detailed production briefs to your AI team.

The AI doesn’t read your mind, it executes your instructions. Feed it vague, low-effort prompts? You’ll get generic, low-effort videos. Give it professional, strategically-aligned briefs? You’ll get videos that are pre-validated to resonate with your target audience.

The Key to Everything

Beyond choosing the right niche and nailing your packaging, the entire system,the strategy, the flywheel, your AI production team, lives or dies by one thing:

The Prompt.

This guide will teach you how to master it, buuut before that… ^^

Phase 1: “Vidrush-Friendly” Niche

Before you write a single prompt, you need to make the most important decision in this entire workflow: picking your niche.

Let’s be clear, this isn’t about following your passion or expressing your creative vision. This is a strategic business decision. Your niche determines everything: your speed, your scale, and ultimately, your profitability.

A “Vidrush-friendly” niche is one that plays to the AI’s strengths while sidestepping its limitations. To qualify, your niche must pass four non-negotiable tests.

1. The B-Roll Availability Test

[!IMPORTANT] This is the golden rule. Break it, and nothing else matters.

Vidrush creates videos by matching your script to visuals from a massive library of existing footage. No relevant footage = no video. It’s that simple.

Before you commit to any niche, ask yourself: “Is there a deep, nearly endless well of B-roll, archival footage, and imagery for this topic?”

If you can’t immediately answer “YES” with confidence, walk away.

Green Flags (Footage Goldmines):

Major Historical Events: World War II is the holy grail here (think P-51 Mustang documentaries, D-Day operations). The Cold War, Vietnam, major revolutions, these all have vast public archives you can tap into endlessly.

Economic & Political Topics: The 2008 financial crisis, Brexit, the Florida housing boom and bust. News agencies have covered these from every conceivable angle.

Well-Known True Crime: Cases like Ted Bundy, the Zodiac Killer, or major heists. You’ve got court footage, news reports, police interviews, documentary recreations, the works.

Science & Nature (The Broad Stuff): Climate change, space exploration (NASA’s entire archive is public domain), wildlife documentaries. These topics have been filmed to death.

Geography & Major Cities: Content about countries, states, or famous cities. Drone footage, street-level video, tourist content, it’s everywhere.

Red Flags (Content Deserts – Run Away):

Hyper-Local Crime Stories: That murder in a small Nebraska town from 1987 with three newspaper photos and a police sketch? The AI will be recycling the same images within two minutes.

Pre-Camera Historical Events: Medieval battles, ancient Rome, prehistoric civilizations. The AI can’t create footage that never existed. You’ll get stuck with the same five paintings on repeat.

Private or Unfilmed Events: Secret corporate meetings, private family dramas, behind-closed-doors negotiations. If cameras weren’t there, you’ve got nothing.

Niche Fiction Without Source Material: Fan theories about obscure anime, original creepypasta stories. No mainstream coverage = no usable footage.

[!TIP] You can check if there is enough Footage with a simple Google Search

2. The Format & Foundation Test

Vidrush was built for one thing: factual, narrative-driven content. Your niche needs to align with this core strength.

What Works (Play to These Strengths):

Documentaries: This is Vidrush’s bread and butter. Historical deep dives, political exposés, social issues, scientific breakthroughs, if BBC or Netflix would make it, you’re golden.

Explainers & Video Essays: Breaking down complex topics (“How the 2008 Crisis Actually Happened,”). Perfect for the AI’s systematic approach.

News Recaps & Analysis: Summarizing current events, analyzing trends, connecting dots. The AI excels at organizing information coherently.

Top 10s and Listicles: Structured, fact-based countdowns work beautifully. The format gives the AI a clear framework to follow.

What Doesn’t Work (The “Don’t Even Try” List):

These aren’t just challenging, they’re technically impossible with current limitations:

Compilations: “Funniest TikTok Fails,” “Best NBA Dunks 2024.” Vidrush can’t grab and aggregate clips while preserving their original audio. It’s not built for that.

Reaction/Commentary: “Reacting to MrBeast’s Latest Video.” No picture-in-picture capability, no way to overlay your commentary on source material.

Software Tutorials: “How to Master Photoshop.” Zero screen recording ability. Dead on arrival.

Vlog-Style Content: The system creates narration-over-B-roll, not personality-driven, face-to-camera content. If you want to be the star, this isn’t your tool.

3. The Repeatable Pattern Test

Your niche needs clear, proven formulas you can replicate. This is crucial for the reverse-engineering process coming in Phase 2.

Before jumping in, study the top channels in your potential niche:

Look for Content Buckets: Do successful channels consistently return to specific themes? A thriving mystery channel might rotate between “Unsolved Disappearances,” “Cold Cases Solved,” and “Paranormal Encounters.” These patterns are gold.

Spot the Title/Thumbnail Formulas: Are there consistent patterns in how winners package their content? If every viral video looks completely different with no connecting thread, you’re looking at chaos, not a system.

If a niche feels random, where every successful video is a complete one-off with no underlying pattern, you’ll be shooting in the dark with every upload. Find niches with blueprints you can follow.

Phase 2: Analysis of a Proven Success

1. Hunt for the “Outlier”

We’re not looking for just any popular video. We’re hunting for an Outlier, a video that’s sending us a crystal-clear signal from the Viewer Patterns that says “This exact combination of topic, angle, title, and structure is hitting different right now.”

What Makes an Outlier?

An outlier is a recent video (last 1-3 months) that pulled 3x to 10x the channel’s typical view count for that period.

Not their all-time best. Not their viral hit from 2021. Their recent breakout.

Why Recent Matters

YouTube optimizes for the individual viewer’s interests, not for global trends. If someone has been watching a specific type of content, YouTube knows they’re currently interested in that topic. That’s why a recent outlier in your niche matters: it means people like your target viewer are actively engaging with that style of content right now.

When you reverse-engineer a fresh outlier, you’re aligning with what the algorithm is most likely to recommend to viewers who already have a watch history signaling interest in that niche. That’s what actually increases your chances of being shown as an impression.

Your Outlier Hunt Process:

1. Pull up a successful competitor in your Vidrush-friendly niche
2. Hit “Videos” → Sort by “Popular”
3. Look at Views and Outlier Score [Or use the NexLev Filter, which is so much better]
4. Spot the spike. There’s a video from 6 weeks ago sitting at 750,000 views with 6x. That’s your outlier.
5. Study the packaging before anything else. Why did this one break out? Compare it to their average:

Step 2.2: Extract the Raw Data

The transcript is your video’s DNA. It contains the exact pacing, vocabulary, sentence structure, and storytelling moves that made this outlier explode. We need to extract it cleanly.

Your Extraction Tool: NexLev Chrome Extension

Forget copying YouTube’s auto-captions by hand. We’re using the NexLev Chrome Extension, built specifically for channel analysis.

1. Install the NexLev Chrome Extension
2. Navigate to your outlier video
3. In NexLev, hit “Transcript” → “Copy Transcript” [Wait a couple min]

Done. You’ve got the raw material.

Pro Move: Build a Pattern Library

Don’t stop at one video.

Find 2-3 top performers from the same content bucket on that channel. If your main outlier is about “Vanishing Hikers,” grab two other successful disappearance videos. Copy those transcripts too.

Why? When you feed multiple winners into the AI in the next step, it can spot the patterns across all of them. You’re not copying one video, you’re extracting a proven formula.

Step 2.3: The Critical Briefing

This is where the magic happens. You’re commanding the AI to reverse-engineer a repeatable success formula.

The Master Meta-Analysis Prompt

Copy this exactly. Every word is intentional:

Analyse the following transcript[s]. I want you to do a meta-analysis and note down the type of **[content/subject]** it covers, the core elements that make this video work, and how we can find similar **[content/subjects]**. Think in terms of talking points, tone of voice, and the angles the story is told from. If we were to find a new subject, what elements does that story need to have to succeed?

Paste your transcript(s) right after this prompt.

Why This Prompt is Surgical:

• “do a meta-analysis…” → Forces the AI to find abstract patterns, not just summarize
• “note down the type of [content/subject]…” → Makes it categorize (deep-dive explainer, emotional narrative, investigative breakdown, educational tutorial, social commentary, etc.)
• “talking points, tone of voice, angles…” → Gives the AI specific lenses to analyze through
• “If we were to find a new subject…” → This transforms analysis into a predictive formula

Your Deliverable: The Success Formula

The AI will hand you back a detailed breakdown, the narrative blueprint of a winner. This is your treasure map.

Here’s what a Success Formula looks like in practice:

Example Success Formula (California Mysteries Channel)

The Hook: Opens with a location-specific chill that sets stakes immediately “Some places in California don’t hide their secrets… after this, you’ll never see California the same way again”

The Pattern of Strangeness: Each mystery shows a recurring, illogical pattern, never just a one-off “It’s not one missing person, it’s dozens… and their footprints don’t fade away, they just end”

Official vs. Unexplained: Creates tension by pitting boring official explanations against specific witness accounts that contradict them Park service says “dehydration” but three witnesses independently describe the same impossible blue light

Sensory Anchors: Grounds each story in specific, unsettling details you can almost feel The sound of a ship’s horn with no ship for miles, ground vibrations that register on no seismograph, a contrail that stops mid-air

Evidence Stacking: Builds credibility with specific, dated sources Ranger Log Entry #447 from 1983, declassified Navy investigation file 7B-Alpha, ham radio recording timestamped 3:17 AM

The Haunting Question: Each segment ends with an unanswerable question that sticks in your brain “The real question isn’t how they disappeared… it’s why the footprints all point to the same tree”

Phase 3: Crafting the Novel Topic

Time to weaponize that Success Formula.

In Phase 2, you extracted the DNA of a winner. Now we’re going to splice that DNA into fresh topics that your audience hasn’t seen before.

Here’s the key: You’re not having a casual chat with AI. You’re directing a research assistant. You’re the strategist; the AI is your researcher. Your job is to guide, push, and refine until you get exactly what you need.

Step 3.1: Your “First Strike” Prompt

Setting Up the Strike:

1. Open a fresh chat with GPT.
2. Paste your entire Success Formula from Phase 2 to give full context
3. Fire off the First prompt immediately after

The Complete First Prompt:

Now I want you to take that information [referring to the Success Formula you just pasted], and find Similar Keypoints/ Storys that follow these same types of narratives.

Ensure that if a hardcore fan watches the video, the points are interesting yet not TOO well known.

DO NOT LOOK AT YOUTUBE VIDEOS. LOOK AT FORUMS AND REDDIT around [Topic]!

Also, find ones that have a good amount of relevant, generic B-roll online if possible.

Why Every Word Matters:

Constraint 1: “Not TOO well known” Without this, the AI defaults to the obvious stuff. Ask for Nevada mysteries? You’ll get Area 51 for the thousandth time. This phrase forces it to dig deeper, finding the gems your hardcore audience hasn’t already watched ten times.

Constraint 2: “DO NOT LOOK AT YOUTUBE” This is crucial. The AI is trained on the entire internet, including every YouTube video transcript. Without this directive, it’ll just regurgitate what’s already viral, trapping you in the content echo chamber. Forums and Reddit? That’s where mostly the untold stories live.

Constraint 3: “Good amount of relevant B-roll” This is your producibility filter. The most fascinating story in the world is worthless if Vidrush can’t visualize it. This constraint pre-screens for ideas that can actually become videos, not just interesting stories.

Step 3.2: Working the Corrective Loop

[!IMPORTANT] The first response is never the final answer.

The AI is brilliant but it needs direction. Think of it like a talented researcher who needs a firm editorial hand. Your job is to push, refine, and redirect until you get gold.

Real Example: The Alaska Mystery Hunt (From Noah’s Actual Session)

Watch how this plays out in practice:

Round 1 - The AI’s First Attempt: The AI came back with five mysteries. Problem? Four of them were about “unexplained sounds.” The AI found a pattern and beat it to death.

The First Correction: Noah didn’t write a new elaborate prompt. He talked to it like he’d talk to a human researcher:

“these are too similar to eachother i want something more variable you did the ones with osunds now i want more variety”.

Round 2 - Getting Warmer: The AI diversified but drifted into pure crime stories and historical events. Lost the paranormal edge that was core to the Success Formula.

The Second Correction: Time to pull it back on track:

“yes but these arent paranomal also use hiker stories but these need to be linked to cryptid or mysteries”

This back-and-forth is THE WORK. This is what turns generic AI output into strategic content gold.

Your Deliverable: The Vetted Content Slate

After this refinement process, you’ll have something powerful:

Each idea on your final list is now:

✅ Strategically Aligned: Matches your proven Success Formula ✅ Fresh: Not the same recycled stories everyone’s seen ✅ Originally Sourced: Pulled from discussions, not other videos ✅ Producible: Pre-screened for B-roll availability ✅ Refined: Shaped by your strategic corrections

The Final Alaska Slate (After Refinement):

Here’s what Noah’s corrective loop produced, five distinct, producible stories ready for Phase 4:

1. Port Chatham “Hairy Man” Attacks – Cryptid terror meets abandoned town
2. Kushtaka of the Tlingit – Shape-shifter folklore with modern hiker encounters
3. Lake Iliamna Monster – Aquatic cryptid linked to missing fishermen
4. Wrangell Mountains “Wild Men” – Humanoid sightings and hiker vanishings
5. Bethel “Little People” Encounters – Indigenous cryptid causing hiker disorientation

Notice the variety? Notice how each fits the formula but tells a different story? That’s not accident, that’s the result of strategic refinement.

The Power Move Most People Miss

Here’s the thing: Most people treat AI like a magic oracle. They ask once and accept whatever comes back. That’s leaving money on the table.

The real power is in the dialogue. In pushing back. In saying “No, that’s not quite right, try again with this angle.”

The AI has access to more information than you could read in a lifetime. Your job isn’t to accept its first answer, it’s to guide that vast knowledge toward exactly what you need.

Next up: Phase 4, where we transform these strategic topics into production-ready prompts that Vidrush can execute flawlessly.

Phase 4: Assembling the Master Prompt

The Four Pillars of a Master Prompt

Pillar 1: 🎥 The Director’s Brief (What the Video Is About)

What it does: This is your elevator pitch. In 2-3 concise sentences, you are telling the AI the core story, the central conflict, and the narrative journey you’re taking the viewer on. This section establishes the primary thesis of your video.

How to nail it: Write as if you’re explaining the video to a friend to get them hooked. What is the big reveal? What is the central tension? Be specific.

My Example (Las Vegas): “This video exposes how even the Mayor of Las Vegas is openly sounding alarms that the city is in crisis. The story frames Vegas as a city caught between collapsing tourism, political fallout, and public fear.”

Why this works: It’s not a generic “video about problems in Vegas.” It establishes a clear protagonist (the Mayor), a central conflict (tourism vs. politics), and high stakes (a city in crisis).

Pillar 2: 🗣️ The Style Guide (How It Sounds)

What it does: This defines your video’s personality and tone. Are you investigating? Exposing? Educating? Warning? This pillar determines the energy of the narration and ensures it matches audience expectations for the niche.

How to nail it: Use descriptive, emotional, and evocative language. Don’t just say “serious”, say “urgent investigative exposé with mounting tension.” Mention specific stylistic devices you want the AI to use, such as cliffhangers, contrasting viewpoints, or a focus on data.

My Example (Las Vegas): “Narrative exposé style — dramatic, urgent, and visual. Anchored by direct mayoral quotes, contrasted with angry Canadian voices. Voiceover feels like an insider breakdown with cliffhangers…”

Why this works: The AI now knows exactly what energy to bring. It understands it needs to create drama, use quotes as evidence, and structure the narrative with suspense.

Pillar 3: 🎯 The Target Audience (Who’s Watching)

What it does: This helps the AI choose the right vocabulary, complexity level, and cultural references to connect with your specific viewer.

How to nail it (The Pro Move): Don’t just describe your audience. Include actual YouTube titles your audience already clicks on. This gives the AI a keyword goldmine and teaches it your audience’s language.

My Example (Las Vegas): “Fans of political/economic exposé YouTube: ‘Vegas is dying,’ ‘America collapsing,’ ‘Tourists reveal truth.’”

Why this works: Those quotation marks are a powerful signal. You’re showing the AI the exact vernacular and click-drivers that resonate with your target demographic.

Pillar 4: 📌 The Core Script (The Meat)

What it does: This is the factual skeleton of your entire video—the information, in the right order, with the right density. This is where most people fail, and where you will succeed. A great outline is not just a list; it is a set of precise instructions for the AI scriptwriter and the AI footage-finder.

The Rules of a Killer Outline

This is where most people fail. They either provide too little information (resulting in a generic video) or they provide the wrong kind of information (resulting in a chaotic, broken video). Your outline is not just a list of ideas; it is a precise set of instructions for both the AI Scriptwriter and the AI Footage Agent. Follow these rules to ensure both are briefed perfectly.

Rule 1: Structure is Everything

The AI reads structure like a roadmap. A clear, logical structure is the single best way to prevent the AI from repeating itself or getting lost.

Use a Hierarchical Format: This is non-negotiable.

• Main points numbered (1, 2, 3…)
• Supporting details bulleted underneath (-)
• Sub-details indented further

Avoid “Fake Depth”: A common mistake is creating an outline with many bullets that are just rewordings of the same idea (e.g., “Potatoes are brown,” “Some potatoes are a brownish color”). The AI Validator flags this internally, as it forces the scriptwriter to become repetitive. Ensure each point introduces a new piece of information or a new angle to create narrative progression.

Rule 2: The Density Formula

You must match your content volume to your chosen video length. This is a direct check against the AI Validator’s EXTREME_CONTENT_MISMATCH warning. An imbalance here is a primary cause of poor-quality videos.

The Golden Ratios:

The Consequences:

• Too few points? The AI will be forced to pad the script, repeating facts and stretching segments, making your video drag.
• Too many points? The AI will rush through each point, resulting in a shallow, incomplete, and unsatisfying video.

Rule 3: Write Visually, Don’t Direct Visually

[!CAUTION] This is the most critical technical rule to understand. The Vidrush Footage Agent is a separate AI that analyzes the final script. It cannot interpret your intent from stage directions.

Giving it direct visual commands will trigger a HYPER_SPECIFIC_VISUALS_IN_SCRIPT warning and produce a poor result.

- ❌ WRONG (Will be ignored or read aloud):
- "In 1992, Jimmy Thompson robbed a bank [SHOW BANK EXTERIOR]. [CUT TO MUGSHOT]"
 
+ ✅ RIGHT (Will pull the right footage):
+ "In 1992, the notorious bank robber Jimmy Thompson orchestrated a brazen daylight heist at the First National Bank in downtown Miami."

Why the second version works: It is loaded with visual keywords that the Footage Agent can understand and search for in its library: “bank robber,” “daylight heist,” “First National Bank,” “downtown Miami.” Your job is not to be a director in the prompt; your job is to be a descriptive writer who provides the AI with the raw material it needs to find great visuals.

Rule 4: Your Script Must Be “TTS-Ready” (Text-to-Speech Ready)

The AI narrator will read EVERYTHING you write—literally. A “dirty” script filled with non-narration text is the fastest way to ruin your audio and will trigger a SCRIPT_NOT_TTS_READY warning.

The Ultimate Cleanup Checklist: Before finalizing your outline, ensure you have removed ALL of the following:

• ❌ Speaker labels (NARRATOR:, VOICEOVER:)
• ❌ Timestamps (0:00-2:30)
• ❌ Stage directions ([PAUSE], [MUSIC SWELLS], [DRAMATIC MUSIC])
• ❌ Production notes ([Note: add dramatic music here], Technobabble:)
• ❌ URLs (or the AI will say “h-t-t-p-colon-slash-slash…”)
• ❌ YouTube auto-caption artifacts ([Music], [Applause], [__])

Your final outline should contain only the exact words meant to be spoken, formatted in a clean, hierarchical structure.

[!TIP] Pro Tip (The AI Cleaner): The fastest way to ensure a perfectly clean script is to use another AI. Before you paste your outline into Vidrush, give it to an AI like Claude with this command:

“Clean this script for a Text-to-Speech engine. Remove all non-narration text, including speaker labels, timestamps, stage directions, and notes.”

My Full Prompt About a Vegas Video:

Here is what a perfect Master Prompt looks like in practice. This is the complete “Las Vegas” prompt, which was generated using the “Direct Replication” workflow. Notice how it flawlessly follows every rule: it has a four-pillar structure, a clean and visual outline, and it avoids all the “Fatal Mistakes.”

This is the level of detail and strategic alignment you should aim for.

🎥 What the Video Is About

This video exposes how even the Mayor of Las Vegas is openly sounding alarms that the city is in crisis. She admits tourism is slumping, Canadians are staying away, and locals feel “nickeled and dimed” by casinos. On top of that, she warns about Nevada being mislabeled as a “sanctuary state,” risking over $100 million in federal funding. The narrative contrasts her cautious, bewildered tone with the raw anger of Canadian commenters who vow never to return to the U.S. The story frames Vegas as a city caught between collapsing tourism, political fallout, and public fear — a downturn so deep that even casino marketing campaigns and desperate fee rollbacks might not be enough.

🗣️ Style of Talking

Narrative exposé style — dramatic, urgent, and visual. Anchored by direct mayoral quotes, contrasted with angry Canadian voices. Voiceover feels like an insider breakdown with cliffhangers (“But Canadians say it’s not just the prices…”). Visual storytelling: empty casino floors, border checkpoints, Canadian destinations thriving, political protest footage. Mix of hard data (tourism numbers, surveys) and personal emotion (comments, lived experiences).

🎯 Who This Video Is For

Viewers concerned about Las Vegas tourism and economy. Travelers debating whether Vegas is still worth visiting. American and Canadian audiences who relate with the content. Fans of political/economic exposé YouTube: “Vegas is dying,” “America collapsing,” “Tourists reveal truth.” Locals who feel the nickel-and-diming culture is ruining Vegas.

📌 Key Facts Covered

1. Mayor’s Admissions

• Tourism slump acknowledged, Canadians singled out as “essential but absent.”
• Genuine confusion over DOJ sanctuary designation: “I don’t understand what the design, I don’t think they understand what they’re designated.”
• Highlights absurdity by listing small towns like Winnamucca, Lovelock, and West Wendover.
• Confesses no DOJ contact: “No, and I’m not sure that a conversation would yield any results.”
• Raises “tremendous fear factor” — even hinting at potential military implications.
• Clarifies Metro Police “work with ICE within the law” but won’t act as ICE.

2. Canadian Reaction

• Comments: “Zero chance of me crossing,” “I’ll never set foot in the U.S. again.”
• Frame as self-preservation: unsafe, unstable, hostile.
• Anger at U.S. culture: xenophobia, guns, fascism, “genocide funding.”
• Redirection of money: Nova Scotia, Quebec, PEI instead of Vegas.
• Some blame Nevada voters: “All 6 electoral votes went to Trump — you brought this on yourselves.”

3. Updated Tourism Stats (Sept 2025)

• 11.3% visitor decline in June 2025 vs. 2024 (3.1M → 2.75M).
• Hotel occupancy down 15%, daily rates slipped 6%.
• Canadian airline traffic plunges: Air Canada -33%, WestJet -31%, Flair -62%.
• Canadian economic contribution: $3.6B in 2024, supporting 43,000 jobs.
• Canadian car visits to the U.S. down 34% in Aug 2025.
• Surveys confirm: 60% of Canadians avoiding U.S., 40% redirecting trips to Canada/Europe.

4. Recent Industry Response

• Some casinos cutting parking/resort fees on select rooms.
• Caesars pushing $300 packages (all-in with taxes, food credits).
• $3.25M international campaign targeting Canadians, Germans, South Koreans.

5. “Trump Slump” Context

• Union leader Ted Pappageorge’s phrase gaining traction — shows decline hits multiple demographics, not just Canadians.
• Framed as political fallout: trade disputes, immigration, hostile culture.
• U.S. projected to lose $12.5B in foreign visitor spending in 2025.

6. Economic Ripple Effects

• Vegas depends on tourism for ~31% of Southern Nevada GDP.
• Small businesses suffer — restaurants, taxis, local shops see fewer customers.
• Wealthy gamblers and mega-events still bring bursts of revenue, but casual tourists are vanishing.
• Locals fear losing $100M+ in federal aid could deepen the crisis further.

The Final Check

Look at that Las Vegas prompt again. Notice what’s NOT there:

• No commands to the AI: It doesn’t start with “Make a video…”
• No stage directions: It doesn’t say [SHOW EMPTY CASINO]. It says “Visual storytelling: empty casino floors…” in the Style Guide and lets the script’s content drive the visuals.
• No timestamps or junk text.

Phase 5: Final Inputs & Execution in Vidrush

Step 5.1: Auto vs. Manual Mode

First decision when you open Vidrush: How much control do you want?

Auto Mode: The Speed Approach

What it is: Paste your prompt, pick a format, hit go. The AI makes all the creative calls. When to use it: When you just want to create content without overthinking and save time.

Manual Mode: The Professional Path

What it is: You stay in the driver’s seat. After pasting your prompt, you’ll guide the AI through specific choices that dramatically impact quality. When to use it: If you want to squeeze every juice out of Vidrush and you think you can tweak the script with this even more.

Step 5.2: The “Stylistic Blueprint”

When you choose Manual Mode, Vidrush asks for a “Reference Video Link.” Most people paste random or Outlier Video in it. That’s not what you should do.

The Human Script Rule

Here’s what Noah discovered through testing: The AI learns to write by studying the reference video’s transcript. Feed it an AI-voiced video? Your script sounds robotic. Feed it a human documentary? Your script sounds human.

It’s that simple. And that powerful.

How to Choose Your Reference Video:

1. Hit YouTube and find a high-quality documentary in a similar style (not necessarily the same topic)
2. Critical requirement: Must be human-narrated. Not AI. Not text-to-speech. Real human.
3. Gold standard channels: “Fern” is explicitly recommended
4. Copy that URL and paste it in the Reference Video field

What the Reference DOESN’T Do:

The reference only affects writing style. It won’t change your visuals, B-roll, or graphics. Those come from your prompt and theme selection.

Think of it like this: The reference teaches the AI how to write. Your prompt tells it what to write about.

Step 5.3: The Follow-Up Q&A

Vidrush will ask you a few questions. Just paste these questions into your active chat in gpt

Don’t overthink this. You’ve already done the thinking. Just transplant your work.

Step 5.4: Theme & Voice

Last two decisions before you pull the trigger. Both matter more than you think.

Theme Selection (Your Visual DNA)

The theme controls fonts, colors, animations, graphics — the entire visual language.

Quick Reference:

• Dark/mysterious topics: Use “Crime” or “History” themes
• News/tech/explainers: “Standard” theme keeps it clean
• Nature/science: “Documentary” theme adds credibility

Match the theme to your content’s mood. A true crime story with the “History” theme? Instant credibility killer.

Voice Selection

Here’s the counterintuitive truth: Perfect isn’t always better.

Noah’s insight: “The goal is to sound less AI.” Sometimes a voice with slight imperfections, what he calls the “Bedroom Recorder” effect actually builds more trust than a polished corporate narrator.

Pro Tips:

• Test different voices. Some are rock-solid, others glitch constantly
• Check the Vidrush Discord for the “Bad Voice List” — avoid known problem voices
• If you get bad audio, don’t edit for hours. Just regenerate with a different voice

The Sweet Spot: Find a voice that sounds like a real person recording in their home studio, not a news anchor or an AI assistant.

Step 5.5: Pull the Trigger

What Happens When You Hit “Continue”:

1. Your video enters the queue
2. Processing takes 50-60 minutes (depending on length and server load)
3. You can close everything, Vidrush runs in the cloud
4. Video appears in “My Videos” when done

That’s it. Go work on your next prompt/ packaging while this one cooks or just take a nap.