고정된 트윗
Ellie Marks
3.9K posts

Ellie Marks 리트윗함

We're giving away the exact list building system behind 23M+ cold emails and 400+ B2B companies. For FREE.
8 phases. 11 video walkthroughs. 2 AI agents that automate most of it.
Here's what's inside:
→ How to define your ICP so it actually filters (not just "Series A SaaS in the US")
→ Multi-source discovery: one provider gives you 50-60% coverage. Two gets you to 80-90%. The math changes everything.
→ Company enrichment and scoring before you ever find a single contact
→ Contact validation: 30-40% of emails in databases are invalid. Skip this step and your sender reputation is toast.
→ Deduplication at company AND people level (most teams skip this and wonder why deliverability tanks)
→ AI personalization built on real data points, not "I saw you work in tech"
→ Activation: getting clean, scored, segmented data out of Clay and into your sequencer
Most teams send raw lists and get 1-2% reply rates. Proper enrichment changes that completely.
To get it:
- Follow @itsalexvacca
- Retweet this post
- Reply "LIST"
I'll DM you the link.

English

signing ecom clients with cold email is a $500K/year PROFIT MACHINE for any agency
took me 3 hours to put together 26 pages of EVERYTHING i know
- scripts that work on brand owners
- lead sources nobody's using
- the BFCM angle that prints
- how to turn replies into calls
after booking 12 calls with ecom owners in 15 days i'm giving it all away
like + comment "ECOM" and i'll send it over
(must follow + RT for priority access)
English

@PatrickSargis That’s sick, AI that’s constantly being trained and updated on what’s working
Felt like I should’ve paid for this sauce
English

This post can add millions to your business - if you apply it.
Everyone knows data-based decisions matter. But how do you actually find the data, and what do you do with it?
We’ve built AI systems trained on winning data that have skyrocketed our creative win rate and cut output time by 70-80%.
Here’s the process:
Once you run an offer for a few months, you have actual market feedback.
What hooks work.
What don't.
What awareness level converts.
What image formats win.
What patterns repeat.
We use Hyros for attribution and Motion for ad data.
Motion's AI chat lets us pull any data we need instantly.
Prompt it to pull your 25 highest-performing creatives and analyze:
Body copy
Headlines
Video scripts
Image concepts
Hook structures
Put each dataset in its own doc.
Drop everything into Claude and ask it to analyze recurring patterns across all winning data points.
You'll find a few principles drive the LARGE majority of results.
For one of our offers:
• 98% of winners had the expert's face in them
• Every top hook led with the core outcome in the first 3-4 words - 100% hit rate
• Contrast-based hooks outperformed everything else
• Out of 50+ image formats we’ve tested - only THREE were big scalers
These patterns are absolute freaking gold.
Then, compile all winning data into a single "creative playbook" per client.
Then build specialized AI prompts/tools:
Headline writer (trained on winning headlines)
Body copy writer (trained on winning copy)
Video script writer (trained on winning scripts)
Image concept creator (trained on winning visuals)
Image ad chief (instant feedback to your designer)
Everything your team creates starts from WINNING DATA and not guesswork.
Two things to keep in mind:
Your AI tools should apply PRINCIPLES - not copy-paste.
We run 70/30 - 70% structure from winning patterns, 30% creative variation.
You still want to have differentiation and unique concepts.
And data decays. The hooks that crushed 90 days ago are fatiguing now.
An ad that spent $40K at $75 CPA six months ago can be LESS valuable than an ad that spent $8K at $80 CPA last week.
Your AI systems need RECENT data, so update your systems weekly.
The flywheel looks like this:
Data trains AI → AI generates creatives → Creatives generate new data → Better data trains better AI.
Every month your win rate goes up.
Every month the system gets smarter.
THAT'S the real moat.
Building compounding intelligence from a system that learns from every dollar you spend.
Most marketers are running new ads based on guesswork.
This is how you achieve real longevity on cold traffic offers and scale them indefinitely.
Back to work 👊
English
Ellie Marks 리트윗함

@elliemdrafts Okay sis! Do you use an automated system or are you manually messaging
English

@AlfieJCarter Is zero message in connection request the move?
English

i sent 1M+ LinkedIn DMs in 2025 using AI LinkedIn outreach
cracked a system that books 3-5 calls per day on complete autopilot
here's the method:
most people send generic "hey saw your profile" messages that get ignored
i use AI to scrape prospects' content and create hyper-personalized audits at scale
the sequence:
- connection request with zero message
- wait 2-3 days
- AI generatesfree value based on their profile
- send it with zero pitch attached
80% reply rate because it feels personal even though it's automated
50% ask what you do
30% book calls immediately
you're positioning as the expert who gives value, not the desperate founder begging for meetings
i documented the entire system in a step-by-step guide
follow, like and comment "AI" and i'll send it to you
English















