Good morning,
Here’s what you’ll find in today’s DTC:
1️⃣ Warm up a new ad account quickly with this clever shortcut
2️⃣ We walk through how to migrate from ChatGPT to Claude without losing any important data
3️⃣ Stop ‘playing with AI’ and start building workflows that make a difference
You’re reading this newsletter along with new subscribers from: Mingle Mocktails, Green Beaver, and Primemax Energy. 👋
[Screenshot from @IstvanicMarin on X]
When you launch new campaigns or take over an ad account, cold starts burn cash.
Meta's algorithm struggles to optimize, and your early margins suffer.
Marin Istvanic shares that resurrecting last year's best-performing assets is the way to go.
Launching historical creative by post ID warms up Meta's algorithm significantly faster than starting completely fresh and allows your team to develop new creative assets.
👉 Try it yourself: Identify last year's highest-performing creatives, grab their exact post IDs, and deploy them in your new campaigns to bypass the cold-start penalty.
🤖 Switching From ChatGPT to Claude to Improve Your Workflow
If you manage multiple brands and rely on AI context to move fast, you've probably thought about switching tools before talking yourself out of it.
Migrating from ChatGPT to Claude is easier than you think.
The thing is, ChatGPT already stores a useful amount of your setup: tone preferences, recurring instructions, project context, behavior patterns.
You can prompt it to output that in a usable format, copy it over, and get most of the way there quickly.
The technical part is straightforward.
The harder part is organizational: files need to be re-uploaded, project context usually needs cleanup, and some instructions written for one tool need to be rewritten for another.
The deeper pattern recognition built from hundreds of chats does not come with you. No export fixes that.
You're moving the blueprint, not the relationship. For most users, the blueprint is enough.
Where Claude’s Cowork has an edge
This becomes more relevant if you're an agency operator repeating the same workflows across multiple clients.
Projects give you persistent context by client, brand, or workstream.
Cowork has Skills: reusable workflows you can run across accounts and use cases.
If you're constantly writing creative briefs, submitting requests, organizing feedback, or building repeatable deliverables, the value here is undeniable.
Build the workflow once. And it’ll save you a lot of valuable time.
Pilothouse shares these helpful migration prompts:
STEP 1: Extract your ChatGPT memories
Paste this into ChatGPT in a normal conversation, not inside a Project:
Output everything you know about me — all stored memories, tone preferences, writing style rules, recurring instructions, projects I’ve mentioned, tools I use, and any behavioral rules I’ve set. Format everything inside a single code block. Use verbatim wording from what you actually have stored. Do not summarize, group, or omit any entries.
STEP 2: Extract context from each ChatGPT Project
Open each Project and paste this into the conversation:
Summarize everything relevant from this project that I should carry forward to a new AI tool. Include: the purpose of this project, any files or documents I’ve shared and their key details, decisions we’ve made, tone and style preferences specific to this project, recurring formats or frameworks we use, and any instructions I’ve given you here. Output inside a single code block. Use verbatim wording where possible. Do not summarize or omit anything.
STEP 3: Inventory your project files
Still inside each Project, paste this:
List every file attached to this project. For each one: the filename, what it contains, and why it matters to this project. I’m migrating to a new tool and need to know exactly what each file is for before I re-upload them.
STEP 4: Rebuild your Projects in Claude
This part is manual:
The takeaway?
Not one-click. Not perfect. But the migration is manageable, and the cleanup process usually reveals that a good portion of what you had wasn't as reusable as it seemed.
AI Is A Stack Of Two-by-Fours. What Are You Building With It?
What do Claude, Motion, and Microsoft Clarity have in common?
They are all AI-powered solutions generating a massive amount of hype and attention.
In this episode, Braydon returns and breaks down how to stop just “playing with AI” and start building actual high-impact marketing systems.
He also identifies the specific tools and custom workflows that manage to drive real brand performance long after the hype cycle dies.
We cover:
Meta takes on Reddit. Forum, a new standalone app, aims to reshape how we handle online group discussions. Read more →
E.l.f. Beauty slashes prices. The beauty brand reverses its tariff price increase to support customers struggling with rising costs. Read more →
DeepSeek triggers AI price war? A 75% price cut on its top-tier V4-Pro model aims to force rapid market adoption. Read more →
Did you find this valuable? Support us by checking out our businesses:
DTC Newsletter is written by Rebecca Knight and Frances Du. Edited by Eric Dyck.
Please note that items in this newsletter marked with * contain sponsored content.
.svg)
.svg)
.svg)
.svg)
.svg)
.svg)
