@mladenovic878 Diagnose before you iterate.
Converts well but has a weak hook rate? Keep the body; swap in hooks from your other winners.
Strong hook, weak conversion? Keep the hook, test a new angle.
We use Reckon (https://t.co/zuWy5sS3Uj) to pull this from the data instead of guessing
@mladenovic878 Research + finding patterns + speed.
Common reasons competitors give users to buy, the ideas they use, common angles, etc. All this research is only useful if you ship the creative ASAP!
Built a competitor ad analyst in Claude.
Been learning performance marketing and creative strategy for 3-4 months. Sharing what I learn, mostly how to automate the ops-heavy parts with AI.
More if you like this.
MCP used: https://t.co/ujNS6NI1fb (works with any MCP client)
@Hsenju Great tip! Reckon does both. Every rival ad is tagged by idea, claim, offer and hook, and 30+ days marks a long-runner. The stronger signal is Scaling: the brand keeps making new cuts of one ad. Copies of the same file don't count, since brands relist ads after every edit.
@fardeentwt Future will be safe only for those who keep on upskilling themselves. You have to be at the top of the game, these people will not just survive, they'll thrive.
I’m not against tools like Cursor; they’re amazing for speed. I just can’t afford to let them think for me at this stage (~2 YOE).
How are you making sure AI isn’t making you dumber?
#AI#Cursor#Devtools#SoftwareEngineering
Are AI tools like Cursor making devs dumber? 🤔
Many devs just generate code, push if it runs, and tweak prompts if it fails without really understanding it.
Here’s how I make sure AI isn’t replacing my thinking:
My simple 2-step system to stay sharp and code faster with AI:
1️⃣ Solve the problem myself, no “give me the answer” prompts directly.
2️⃣ Write the algorithm in plain English → let Cursor implement → review & optimize