Meta ads prompts for Claude, ChatGPT & any AI agent

Pull Acme's last 14 days at campaign, ad-set and ad level, then tell me whether CAC is up because of the learning phase, audience saturation, creative fatigue, or auction overlap, and what you'd change. Don't touch anything yet.

Is the 'Prospecting – US 1% LAL' ad set really fatigued? Check frequency, the 7-day CPM trend, and the quality / engagement / conversion rankings. Only call it fatigued if 3+ days are trending the wrong way.

My purchase campaign keeps showing 'Learning Limited.' Tell me why, and whether consolidating these 3 overlapping ad sets or broadening the audience would get it to about 50 purchases in 7 days.

Rank every active ad set by cost per purchase over the last 14 days, pause the bottom two, and move their daily budget to the top-ROAS ad set.

Launch a Sales campaign with Advantage+ targeting in the US, $40/day, optimized for purchases. Build 10 ads from the images I uploaded to Xylo, with 5 primary texts and 5 headlines each, a Shop Now button to acme.com/sale, product extensions using my all-products set, and this account's Advantage+ creative defaults and UTMs. Keep everything paused for review.

Duplicate my top ad set into a new 2% lookalike of 180-day purchasers, keep the same creative, start it at $30/day, and exclude anyone who bought in the last 30 days.

Create a budget schedule that raises this campaign's daily budget 50% from Friday 6pm to Sunday 11:59pm for the weekend sale.

Build an Advantage+ catalog ad from my all-products set with a Shop Now CTA, dynamic formats, and this account's creative defaults.

I uploaded 6 videos to Xylo. Pair each Feed cut with its Story/Reel version, name the groups by product, and push them to Meta so they're ready to use in ads.

Make one flexible ad from these 4 images and 2 videos so Meta picks the best combination, with 5 primary texts, 5 headlines, and a Learn More CTA to this link.

Build a single multi-media ad that shows these 3 images and 2 videos together in one ad, with 5 primary texts, 5 headlines, and a Shop Now CTA to this link.

Swap this new primary text into the 'Spring – Hook B' ad. Keep it a single-image ad, don't convert it to dynamic creative.

Analyze my best ad against its real delivery: hook rate, 3-second and 15-second hold, and the full drop-off curve. Tell me what's carrying it and the one thing to change.

Decide whether the 'UGC – Testimonial' ad should be scaled, iterated or killed based on its delivery, then rebuild the hook on a different framework. Try Cost Narration or RSV, and give me the new script.

Give me 5 new 15-second video hook ideas for this product, each on a different framework, with the on-screen text and spoken first line for each.

Rewrite this script to open with a Zeigarnik open loop, hit value in under 3 seconds, and keep every product claim truthful and substantiable.

Build a 1% US lookalike from my 180-day purchasers, plus a CRM custom audience from the customer list I'll paste, and create an exclusion of everyone who bought in the last 30 days.

Search detailed-targeting interests for 'home fitness,' show me the 10 largest, and give a delivery estimate for a $50/day purchase-optimized ad set in the US.

Draft this week's client report: top 5 and bottom 5 ads by ROAS, the root cause behind any big swing, and a next-week plan. Compare the last 7 days to the 7 before.

Break down this month's spend and ROAS by primary text and headline so I can see which copy is actually winning.

Query the warehouse for daily spend, purchases and ROAS by campaign over the last 30 days, and flag any day a campaign moved more than 30%.

Publish this reel to our Instagram with this caption, turn comments off, and confirm we're under today's publishing limit first.

Pull the newest comments on our top ad, reply to the buying questions with this answer, and hide anything spammy or abusive.

Send yesterday's website purchases to the pixel via CAPI. Include event IDs and order values so they deduplicate against the browser pixel.

Pull every Facebook and Instagram ad everydaycoffee.com is running from Meta's public ad library, group them by angle, and tell me which offers and hooks they lean on hardest.

Search Meta's ad library for 'cold brew' and show me how the category is being advertised: the common hooks, offers, and formats.