--- name: al-599-get-fresh-ad-ideas-sent-to-you-every-morning-for-approval description: "Perplexity's customer story on Get Maine Lobster, a US seafood delivery company founded by Mark Murrell, describes a pilot of Perplexity's agent platform Computer as a marketing 'command center'. Every morning it analyzes Meta and Shopify data, learns the brand's value propositions and visual style, and studies existing Meta ads across ecommerce, not just direct competitors, for campaign ideas. Through the day it generates ad ideas, writes creative briefs and makes static ads, sending them to Mark in Slack for review, and connects with Higgsfield AI for video ads. Run side by side with the brand's usual non-AI ads, the AI campaign got 63% higher return on ad spend while spending 2.1x as much, bringing in 3.5x the revenue, and micro-testing creative variations (avatars, outfits, settings, colours) found ads with 14% lower acquisition cost. When orders outstrip lobster supply, it also updates arrival dates to when enough stock will be in. These are one company's pilot results, reported…" --- # Get fresh ad ideas sent to you every morning for approval: an AI agent reads yesterday's ad and store data, studies ads from other brands, writes creative briefs and drafts static ads, and sends them to Slack for your yes or no before anything runs Perplexity's customer story on Get Maine Lobster, a US seafood delivery company founded by Mark Murrell, describes a pilot of Perplexity's agent platform Computer as a marketing 'command center'. Every morning it analyzes Meta and Shopify data, learns the brand's value propositions and visual style, and studies existing Meta ads across ecommerce, not just direct competitors, for campaign ideas. Through the day it generates ad ideas, writes creative briefs and makes static ads, sending them to Mark in Slack for review, and connects with Higgsfield AI for video ads. Run side by side with the brand's usual non-AI ads, the AI campaign got 63% higher return on ad spend while spending 2.1x as much, bringing in 3.5x the revenue, and micro-testing creative variations (avatars, outfits, settings, colours) found ads with 14% lower acquisition cost. When orders outstrip lobster supply, it also updates arrival dates to when enough stock will be in. These are one company's pilot results, reported by the vendor. ## Inputs to collect first Ask the user for anything below that they haven't already given you. Don't guess missing facts. - Or describe them ## Instructions Do the task below with the user's inputs in place of each [bracket]. 1. Each morning, export yesterday's ad results (spend, clicks, sales or leads per ad) and your top-selling products or listings. 2. Ask your AI assistant: "Here is yesterday's data and a one-page note on our brand (what we sell, why people buy, our look and tone). Which ads are working and why? Then look at 5-10 ads from other businesses selling to similar customers [paste or describe them] and suggest 5 new ad ideas." 3. For the 2 best ideas: "Write a short creative brief for each (audience, hook, offer, visual, call to action) and 3 versions that change only one thing (setting, person, colour or headline)." 4. Make the images in your design tool, post them to a Slack channel or your notes, and approve or reject each yourself. Nothing goes live without your yes. 5. Run approved versions as small tests against your usual ads and compare cost per sale or lead after a week. 6. Keep claims honest: no prices, availability or results you can't back up. ## Safety checks (share these with the user) - Check every figure, price and total against your own records before anything is sent, posted or booked. A person approves refunds, discounts, payments and price changes. - Remove names, contact details, account numbers and other personal data the task doesn't need before you paste, and check your AI tool's data and training settings. - Read and edit the draft before it reaches a customer or goes public. If you later automate this prompt, start with "draft and notify me", not "send automatically". - Check names, dates, numbers and any claim against your own source material before you use the answer. AI can misread or invent details. - Say clearly which parts need a person to review or approve before anything is sent, paid, posted or filed. ## About this skill - Works with: Perplexity Computer or ChatGPT/Claude, Meta Ads Manager, Shopify or your sales data, Slack, an image or video tool - Based on this real-world example: https://www.perplexity.ai/hub/customers/get-maine-lobster - Free prompt from AgentPlaybook Real Estate: https://agentic-library.netlify.app/prompt/al-599-get-fresh-ad-ideas-sent-to-you-every-morning-for-approval-an/ - To install in an agent that supports Agent Skills, save this file as SKILL.md in a folder with the same name as `name` above (for example in your skills directory). Install steps for Claude.ai, Claude Code, Cursor and Codex: https://agentic-library.netlify.app/skills/