- Bionic Business
- Posts
- Issue #83: How To Use AI Video Clones To Sell
Issue #83: How To Use AI Video Clones To Sell

Good morning.
A software sales team just shared their numbers with me:
They went from a 4% response rate to 52%. Same list. Same offer. Same quarter.
How? They stopped sending text emails and started sending personalized AI videos.
They only recorded one video.
Two minutes. One take. That single recording became 1,000 personalized videos.
Each prospect heard their name, saw their company website in the background, got specific references to their industry challenges.
The AI models and system handled all of it.
This isn't some venture-backed startup with a massive tech stack. It's a 5-person sales team using $29/month software.
Record once, personalize infinitely.
Two minutes of human effort generating thousands of unique human touchpoints.
A 30% increase in qualified pipeline with less work than writing a single email sequence.
While everyone's debating whether AI will replace sales processes and salespeople, smart entrepreneurs are using it to multiply themselves.
You don't scale by doing more. You scale by recording once and letting the machine do the rest.
Let me show you exactly how to build this system.
— Sam
IN TODAY’S ISSUE 🤖

The Master Template Formula (What to Record, What to Leave Blank)
The 3-Layer Personalization Architecture
Platform Breakdown: Tavus vs. HeyGen vs. Synthesia (With Real Pricing)
The CSV-to-CRM Automation Pipeline
Why 52% Click-Through Isn't Even the Ceiling
Let’s get into it.

Free Cortex Preview
If you’re a free subscriber, this email is a good example of the kind of strategies and methods you get inside Cortex, which is the premium membership version of Bionic Business.
In addition to 3-4 full, unrestricted regular issues (like this one) per month, you also get:
1x SIGNALS Strategic Briefing issue (strategic moves you must make now to secure your business revenue, market share, and profits).
1x CIRCUITS Tactical Guide issue (workflows, how-to’s, prompts, very practical implementation inside a self-driving business).
For free subscribers, you’d get a redacted and summarized version of this email (and same for other issues, only Cortex subscribers get the full emails loaded with the good stuff).
But right now, as Cortex is open for a few days (and it only opens once per month), you get access to this full issue.
So, if you’d like to receive full issues like this one, and the monthly special Signals and Circuits issues, then you should sign up for Cortex, which is $50/month (and you have full archives of previous regular issues for as long as you’re a member).
(By the way, you only get the issues for the month you’re subscribed to and not the Signals and Circuits archives, so if you wait another month, you’ll miss out on past issues).
Also, almost all issues on the website are only accessible with a Cortex subscription.
So, if you want full, unrestricted access to those regular issues, that’s only available when you:
Enjoy the full Cortex issue below, covering prompts for producing high-converting ads—not just for Facebook but any platform, including organic social posts, too (just tweak the prompts).

The Master Template Formula (What to Record, What to Leave Blank)

Personalized video outreach isn't new.
What's new is that you don't have to record 1,000 videos anymore.
The shift happened quietly. While everyone was obsessing over ChatGPT, the video AI platforms solved personalization at scale.
Not through better cameras or editing software. But through templates and variables.
One recording becomes infinite variations.
This changes the entire economics of sales outreach and inbound sales videos.
The constraint was always time. Record 100 personalized videos? That's two full days. Nobody did it. The juice wasn't worth the squeeze.
Now? Two minutes. One template. Unlimited personalization.
The formula is deceptively simple:
You record. AI clones.
Your: The core value prop, the social proof, the call to action. The parts that never change.
The AI model: Their name, their company, their website in the background, their specific pain points, their industry jargon.
Most people screw this up when they try to personalize everything. The greeting, the middle, the close, every transition. The video becomes uncanny valley territory. Too many variables. Too many seams showing.
The best templates are 80% static, 20% dynamic.
That 20% is what makes it feel personal. The 80% is what makes it scalable.
Think about it: When you're on a sales call, how much is truly personalized? You mention their company, their role, maybe a specific challenge. The rest? It's your standard pitch. The same social proof. The same value prop. The same close.
The template system just admits what's already true: most of your pitch is repeatable.
The key is accepting that personalization is about a few key moments, not the entire conversation.
Master that ratio, and you've cracked the code.
Now, the question is:
Should you use AI to clone a video of yourself into hundreds of variations, some with personalization?
Whether you do a cold outreach or send traffic to multiple versions (clones) of one single video, only you can decide whether you want to do this or not.
Would your market and audience lose trust in you if they knew they’re watching an AI clone?
Or are they more concerned about feeling understood and offered value, regardless whether it happens to be a copy of you speaking to them (if they can even tell)?
You’re not removing what’s personal. It’s still you, on a screen. Some parts of what you do will be you recording multiple versions. The rest is done with technology, that happens to be AI video models.
Still, some of this might be controversial.
I’m going to show you how to do this easily. Whether you do this or not depends on you and your business, and your market.
The 3-Layer Personalization Architecture

To understand why this works, you need to see the system as three distinct layers. Each layer has a job. Mix them up, and the whole thing falls apart.
Layer 1: The Base Recording (Your Two Minutes)
This is your foundation. The skeleton that everything builds on.
Record it like you're talking to your ideal prospect. Not a specific person but the archetype. Use "you" language. Keep it conversational.
Hit these marks:
0:00-0:15 - The Hook "I noticed you're..." (Leave a pause after "noticed you’re"—AI will insert their actual situation)
0:15-1:00 - The Problem/Solution Bridge Your core value prop. What you do. Why it matters. This never changes.
1:00-1:30 - Social Proof "Just like when we helped [similar company] achieve [result]..." Generic enough to apply broadly, specific enough to feel relevant.
1:30-1:45 - The Ask "I'd love to show you exactly how this would work for..." (Pause for company name insertion)
1:45-2:00 - Next Steps "I have some time on [day] if you're interested..." (AI customizes the actual day based on send date)
The magic is in the pauses. Those half-second gaps where AI drops in personalization. Too many pauses, it feels choppy. Too few, it feels generic.
Layer 2: The Data Layer (Your Prospect Intelligence)
This is where your CRM data becomes video fuel.
Every row in your CSV is a complete personalization package:
Identity Data: Name, company, title
Context Data: Industry, company size, growth stage
Pain Point Data: Their specific challenge (pulled from intent data or discovery)
Timing Data: When to suggest a meeting
Visual Data: Company URL for background display
The platforms can now pull from 20+ data fields. But more isn't better. The best performing videos use 3-5 personalization points. Enough to feel custom, not so much that it feels creepy.
Layer 3: The Synthesis Layer (Where AI Does Its Magic)
The platform takes your base recording, maps it to the data fields, and generates unique outputs.
What actually happens:
Voice synthesis modulates your tone to pronounce names naturally
Lip sync adjusts to match the new audio
Background replacement shows their actual website
Text overlays add their company logo or name
Timeline adjustment keeps everything synchronized
The computation happens in 2-5 minutes per video. Not instant, but fast enough to generate hundreds overnight.
Is this a deepfake? I’m not sure it is. It's not replacing you with an avatar. It's multiplying your actual presence.
You record once in your voice, with your face, in your style. The AI just handles the tedious parts: saying 1,000 names, referencing 1,000 companies, customizing 1,000 calls to action.
It's the difference between a form letter and a mail merge. Same efficiency gain, but for video.
Three layers. One system. Infinite scale.
And the whole thing runs on software that costs less than your monthly coffee budget.
Platform Breakdown: Tavus vs. HeyGen vs. Synthesia (With Real Pricing)
Let's cut through the marketing BS and look at what actually works.
I've tested all three. My clients have burned through credits on all three. Here's the real breakdown.
Tavus: The Ferrari (When You Need 1,000+ Videos/Month)
What they're good for: True dynamic personalization at massive scale. Upload once, generate thousands. They're built for this exact use case.
The workflow: Record your 2-minute template → Upload CSV with 10,000 rows → Walk away. Come back to 10,000 personalized videos with different backgrounds, names, company references, everything.
Real pricing: They don't publish it. It's "contact us." From what I've seen, expect $500-2,000/month depending on volume. You're paying for the ability to generate unlimited variations without credits.
The catch: Minimum commitments. Enterprise-style contracts. If you're doing less than 500 videos/month, you're overpaying.
Who should use it: Sales teams with 10+ SDRs, agencies running campaigns for multiple clients, anyone sending 1,000+ videos monthly.
HeyGen: The Toyota Camry (The Practical Choice)
What they're good for: Balance of features and price. Good personalization, decent speed, Salesforce integration out of the box.
The workflow: Similar to Tavus but with credit limits. Record template → Connect to CRM → Set triggers for automatic generation. More hands-on than Tavus.
Real pricing (at the time of publishing):
Starter: $29/month (not enough for this use case)
Business: $89/month (includes personalization features)
Credits for personalization: ~$0.10-0.50 per video depending on length
The catch: Users report quality degradation recently. Generation times slowing down. The platform feels like it's struggling under growth. Support is getting worse.
Who should use it: Entrepreneurs, business owners, marketers, or teams sending 50-500 videos/month who want CRM integration without custom development.
Synthesia: The Honda Civic (Reliable, Limited)
What they're good for: Rock-solid platform. Great for training videos, product demos, corporate communications. Less dynamic for sales personalization.
The workflow: More manual. Upload your script → Choose avatar → Generate video. Personalization exists but it's clunky compared to the others.
Real pricing (at the time of publishing):
Starter: $29/month for 10 minutes
Creator: $89/month for 30 minutes
Custom avatars: $1,000/year extra
The catch: It's built for avatar videos, not personalized outreach. You can hack it for sales videos, but that's not its sweet spot.
Who should use it: Entrepreneurs, business owners, marketers, etc. that also need training videos, product demos, and other corporate content. Not ideal for pure sales outreach.
The Dark Horse Options
Reachout AI: Purpose-built for sales videos. ~$39/month. Fewer features but laser-focused on this use case. If you just want personalized prospecting videos, this might be your play.
BombBomb/Vidyard: Old school but reliable. More expensive (~$100/month) but mature platforms with proven deliverability.
The Integration Reality
Here's what the platforms don't tell you:
API availability:
Tavus: Full API, build whatever you want
HeyGen: API in beta, Zapier integration works
Synthesia: API exists but it's enterprise-only
CRM connections:
HeyGen: Native Salesforce, HubSpot coming
Others: Zapier or custom development
Email platform compatibility:
All work with major platforms (Outreach, SalesLoft, HubSpot)
Video hosting is handled by the platform
Thumbnail generation automatic
My Recommendation
Start with HeyGen at $89/month. Test the workflow. Prove the ROI.
Once you're sending 200+ videos monthly and seeing results, evaluate:
Staying with HeyGen if it's working
Upgrading to Tavus for true scale
Moving to Reachout AI for simplicity
Don't start with Synthesia for sales videos. It's like buying a pickup truck to race Formula 1. Great tool, wrong job.
The platform matters less than the process. Pick one, commit for 30 days, optimize relentlessly.
Tools don't create results. Systems do.
The CSV-to-CRM Automation Pipeline

Forget everything you think you know about complex integrations. This pipeline is simpler than setting up a Zap.
The whole system runs on three connections. That's it.
The Data Flow (What Actually Happens)
Your CRM holds the truth. Everything flows from there.
Step 1: The Trigger
New lead scores above 70
Demo request submitted
Prospect opens 3+ emails
Manual flag by SDR
Step 2: The Data Package CRM exports these fields:
First_Name: John
Company: TechCorp
Website: techcorp.com
Title: VP Sales
Pain_Point: "scaling team from 10 to 50"
Industry: SaaS
Last_Touch: Downloaded pricing guide
Meeting_Day: Tuesday (calculated from today + 2)
Step 3: The API Call Your automation platform (Zapier, Make, or native integration) sends this package to your video platform. One POST request.
Step 4: The Generation Platform receives data → Maps to template variables → Generates video → Returns URL
Step 5: The Injection Video URL gets pushed back to CRM → Inserted into email template → Sequence continues
Total time is about 3-5 minutes from trigger to sendable video.
The Three Ways to Build This
Option A: The No-Code Path (Zapier/Make)
Cost: $20-50/month
Setup time: 2 hours
Technical skill needed: None
The Zap:
Trigger: New CRM record with tag "send_video"
Action: Create row in Google Sheets (backup)
Action: Send to video platform via webhook
Delay: Wait 5 minutes
Action: Update CRM with video URL
Action: Trigger email sequence
This works for 90% of teams. It's not elegant, but it works.
Option B: Native Integration (HeyGen + Salesforce)
Cost: Included with HeyGen Business
Setup time: 30 minutes
Technical skill needed: Basic sales CRM admin (like Salesforce, etc.)
Direct connection. No middleware. HeyGen reads from Salesforce, writes back to Salesforce.
The limitation: You're locked into their workflow. Less flexibility, but less that can break.
Option C: Custom API Pipeline
Cost: Developer time
Setup time: 1-2 weeks
Technical skill needed: Python/JavaScript
Build it exactly how you want. Rate limiting, error handling, bulk processing, custom logic.
Only worth it if you're sending 1,000+ videos monthly.
The Gotchas Nobody Mentions
Rate Limits: Most platforms cap at 100 videos/hour. Your beautiful automation hits a wall at scale. Solution: Queue management. Spread generation across hours, not minutes.
Name Pronunciation: "Wojciech" doesn't sound like it looks. Add a phonetic column to your CSV:
First_Name: Wojciech
Phonetic: VOY-check
Failed Generations: 3-5% of videos fail. Network timeouts, API errors, whatever. Build retry logic. After 3 fails, default to the generic version.
The LinkedIn Problem: LinkedIn doesn't embed videos. You need landing pages. The platforms provide hosted URLs, but they look generic. Better solution: iframe embed on your domain.
The Advanced Pipeline (What We Actually Built)
Here's the system running at one of my clients:
Morning Batch (6 AM)
Cron job triggers
Pulls all MQLs from last 24 hours
Enriches with Clearbit/Apollo data
Generates videos in priority order
High-value accounts get custom templates
Continuous Stream (Business Hours)
Webhook listens for demo requests
Instant trigger to video generation
5-minute SLA from request to video email
SDR notified when sent
Evening Reconciliation (6 PM)
Check all videos generated successfully
Retry failures
Update CRM with engagement data
Flag videos with >80% watch time for follow-up
That all sounds good. But it wouldn’t be complete without a data enrichment layer.
The Data Enrichment Layer
Don't just use what's in your CRM. Enhance it:
LinkedIn scrape → Recent posts, job changes, connections
Company news API → Recent funding, announcements
Tech stack tools → What software they use
Intent data → What they're researching
Feed all this into your personalization. Mention their recent Series B. Reference their tech stack. Comment on their LinkedIn post from yesterday.
Don't build the perfect pipeline on day one. Build the minimum viable pipeline, then optimize based on actual results.
Start ugly. Optimize later.
Why 52% Click-Through Isn't Even the Ceiling
That 52% CTR sounds impossible until you understand what's actually happening.
Traditional email: Your prospect skims text, maybe reads the first line, deletes it. Seven seconds of attention if you're lucky.
Video on landing page, blog post, or via email: They see their name in the thumbnail. Their company website in the background. They think, "Wait, did this person really record a video just for me?"
They click. Of course they click.
Let’s talk about why they do.
The Psychological Triggers That Actually Drive Response
The Reciprocity Bomb
When someone believes you spent personal time on them, they feel obligated to reciprocate. Two minutes of video feels like 10 minutes of effort. The guilt of not responding becomes overwhelming.
One client team added this line: "I spent a few minutes reviewing your site and recorded this quick video with some specific ideas for TechCorp..."
Response rate jumped 15%.
They didn't spend more time. They just made the perceived effort explicit.
The Pattern Interrupt
Your prospect gets 50 emails a day. All text. All the same. Then suddenly they see a video thumbnail with their company logo visible.
A/B test data shows thumbnails with the recipient's website visible get 3x higher click rates than generic video thumbnails.
The Effort Heuristic
Humans use effort as a proxy for value. High effort = must be important.
Text email = zero effort in their mind
Generic video = some effort
Personalized video = high effort
Personalized video with specific details = maximum effort
Each level multiplies response rates.
The Optimization Plays That Push Past 60%
The Screenshot Hack
Instead of just showing their website, screenshot a specific page. Their pricing page. Their about page. Their recent blog post.
"I was looking at your pricing page and noticed..."
One team using this got 67% response rate.
The LinkedIn Stalk
Reference something from their LinkedIn. Not just their company. To them personally.
"Saw you just celebrated 2 years at TechCorp..."
"Noticed you're hiring for SDRs right now..."
"Your post about scaling challenges resonated..."
This data is scrapeable. Automatable. But it feels personal.
The Time Bomb
Add urgency without being salesy.
"I'm doing a few of these videos this week for companies in your space..." "Wanted to share this before our pricing changes next month..." "Recording these for a few VPs dealing with [specific challenge]..."
Limited time + limited availability + specific relevance = response.
The 80% Response Rate Unicorns
Yes, they exist. Here's how:
Trigger-Based Sending
Don't batch and blast. Send based on signals:
They just raised funding (12 hours after announcement)
They posted a job for a role you help with (same day)
Their competitor just became your customer (next morning)
They downloaded your content (within 1 hour)
The fresher the trigger, the higher the response.
The Double-Tap Method
Video 1: Sent via email
Video 2: Sent via LinkedIn 24 hours later
Video 3: Text message with video link 48 hours later
Different message each time. Same personalized approach.
One B2B team using this triple-touch: 78% response rate. Not open rate. Response rate.
The Executive Assistant Bypass
For C-suite targets, send the video to their EA with this message:
"I recorded a brief video for [Executive Name] about [specific challenge]. Could you share it with them if you think it's relevant? The video is 90 seconds."
EAs forward videos. They don't forward cold emails.
The success rate is about 45% get forwarded. And 80% of forwarded videos get responses.
Why Most Teams Stop at 20%
They make these mistakes:
Over-personalizing the wrong things: Saying their name 5 times doesn't help. Mentioning one specific, relevant business insight does.
Under-personalizing the thumbnail: The thumbnail is 80% of the battle. If it looks generic, you've lost before they click.
Sending to everyone: The highest response rates come from surgical targeting. 50 perfect prospects beat 5,000 mediocre ones.
Giving up after one video: The money is in the sequence. Video 1 gets 40%. Video 2 gets another 20%. Video 3 gets another 15%.
The Uncomfortable Truth
Most sales teams are terrible at personalization.
They think personalization means using someone's name. That's not personalization. That's mail merge from 1995.
Real personalization means:
Understanding their actual business situation
Referencing specific, timely challenges
Showing you've done homework
Providing relevant social proof
Making a specific, valuable offer
AI video just makes real personalization scalable. But if your message sucks, your response rate will too.
The entrepreneurs and teams hitting 70%+ response rates were already good at sales. They just found a way to multiply themselves.
52% is what happens when average salespeople use great technology.
70%+ is what happens when great salespeople use great technology.
Humans are the X factor.

You wake up. Check your CRM. Ten new videos were generated overnight. Five responses waiting. Two meetings already booked.
You recorded nothing. The system runs itself.
That's the goal. One month from today, you're running a machine that multiplies your presence while you sleep.
One template. Ten videos. Manual sends.
Start building your pipeline today. Not next week. Not after the holidays. Today.
While your competitors are still debating whether AI video "feels authentic," you'll be booking 20+ meetings from a single recording session.
This is about earning more opportunities to have human conversations.
The video gets you in the door. Your expertise—you being a human being—closes the deal.
This puts AI in service of you having more personal, human conversations.
Until next time,
Sam Woods
The Editor
.