Sheff

[Sheff @sheffnyc]

@alexgoughcooper Would love to check it out

[Sheff @sheffnyc]

@alexgoughcooper Parker

[Alex Cooper @alexgoughcooper]

Introducing Parker: The world's first AI Creative Director.

100+ brands like Grüns, Lume and Legends use us already.

RT + Comment "Parker" and I'll send you 100+ AI prompt engineers for hire.

[Sheff @sheffnyc]

@Aazarshad Video

[Sheff @sheffnyc]

@lorenzo_pravata Formats

[Sheff @sheffnyc]

@oliverwhudson Playbook

[Olly Hudson @oliverwhudson]

We’ve built a Meta creative playbook, (built from the learnings behind £400M in revenue💰) and you can download it for free.

Meta's Andromeda now only cares about how much quality creative you can give it and how fast.

→ It can test 5,000 ads a week.
→ It learns in real-time.
→ It rewards creative variation and rapid iteration.

If you're still running 3-5 ads per month and hoping for scale... you're feeding a machine that runs on volume with crumbs.

That’s why we built a system designed specifically for this new Meta reality and now we’re giving it away.

🚨 The 2025 Creative Testing Playbook

What’s inside:

✅ Testing architectures for Persona, Concept, and Sprint isolation (£15k–£100k+ budgets)

✅ Modular frameworks to script, shoot, and remix hooks, bodies & CTAs at scale

✅ 30-day optimisation cadence with KPI guardrails and budget distribution cheatsheets

✅ Plug-and-play checklists for reviews, setup, and analysis

✅ The 7 mistakes that silently kill ad accounts and how to avoid every one

It’s the exact process we use with brands spending 7 and 8 figures a year.

Want it?

Retweet this post
Comment “PLAYBOOK” below and I’ll send it your way.
(Make sure you're following so I can DM you).

[Sheff @sheffnyc]

@adswithani Hypothesis

[Sheff @sheffnyc]

@tyler_agg Reddit

[Sheff @sheffnyc]

@oliverwhudson Script

[Olly Hudson @oliverwhudson]

This prompt is behind 80% of our top-performing TOF scripts.

🔥The feedback from our most recent podcast episode has been next level.

In the episode, Loukas shared a prompt that tackles one of the biggest challenges creative strategists face: going beyond surface-level pain points.

Most ad scripts mention what the product does (clearer skin, better sleep, more energy) but they miss the deeper emotional drivers behind why people actually buy.

That’s where the Human Desires Prompt comes in.

It’s a tool we use internally to help strategists:

- Link product benefits to emotional drivers like confidence, identity, or belonging
- Write stronger stories in the unaware and problem aware stages of the funnel
- Uncover the “why behind the buy” to make scripts feel more human and relatable

Loukas uses this exact prompt to review scripts and spot emotional gaps, especially in founder stories, testimonials, and top-of-funnel performance creative where trust and connection matter most.

It’s a core part of our internal training and has consistently helped improve creative quality across client accounts.

Want the prompt from the ep?

Retweet this post

Comment “script” and I’ll DM you.

[Sheff @sheffnyc]

@THArrowOfApollo JULY

[Sheff @sheffnyc]

@lorenzo_pravata VSL

[Sheff @sheffnyc]

@Aymenzayani_ CS

[Sheff @sheffnyc]

@sourfraser Workflow

[Aryan Mahajan @aryanXmahajan]

this might be the STUPIDEST thing i've ever done...

leaking my ENTIRE automation library

here's what you get:
- 3,000+ n8n workflows including AI ad creative agents
- AI UGC system that saves $247 per video (5-minute creation)
- 100+ AI prompts that don't suck
- Claude MCP AI agents for LinkedIn automation
- AI prompts that write better than most humans
- AI Viral Content OS (complete content empire)

basically everything you need to automate your business

Like + comment “VAULT” + repost (must be following so I can DM you)

PS: only FREE for 24hrs before it goes into internal client systems

[Sheff @sheffnyc]

@aryanXmahajan Vault

[Sheff @sheffnyc]

@JoeJMarston n8n

[Joe Marston | Ecom Growth @JoeJMarston]

We’re using n8n to turn hours of manual research into ad strategy in under 10 minutes.

Scraping Reddit threads, customer reviews, support tickets, surveys… then piping it into AI workflows that generate hooks, angles, and even static mockups.

We automate the part no one wants to do and feed Meta exactly what it needs to scale.

I broke this down at the @gorgiasio Human vs AI event and most of the room hadn’t even heard of it.

Meta now uses AI (andromeda) to deliver content based on user behaviour, at the individual level.

You prefer short-form statics. I prefer long-form founder stories. Meta serves us both based on 15+ years of behavioural data.

That means creative is now your targeting.

So if you’re still pushing five variations of the same concept, you’ll fall.

To compete, you need:
✅ High volume of content
✅ Multiple creative formats
✅ Messaging that speaks to specific buying triggers

This is what we’ve built internally.

Here’s what we’re doing across our DTC client base:

1. Scraping real customer feedback
From reviews, Reddit, email support, post-purchase surveys, any voice-of-customer data we can find.

2. Running it through AI agents built in n8n
This no-code workflow lets us sort that data into categories: benefits, objections, pain points, buying triggers.

3. Turning that into creative strategy
We generate ad concepts, angles, static mockups, and UGC briefs, all based on what customers are actually saying.

It’s fast. Scalable. And backed by real insight.

We’ve used this exact process to build our top-performing ads across hundreds of brands.

And as Meta shifts further toward AI-led delivery, this kind of system is only going to become more important.

We’re not guessing. We’re using customer data, n8n, and AI to drive creative decisions, daily.

Want to know how? I ran through this last weeks at the Gorgias AI vs Human event. Want the recording?

🔄 Retweet this post
💬 Comment below “n8n” and I’ll send you the recording.

(make sure we’re following so I can DM you).

[Sheff @sheffnyc]

@jspujji ICP