starting a new agency tbh

Most SaaS founders want consistent qualified demo calls but are only sending 2,000 cold emails per month. The math doesn't make sense. If your email-to-call ratio is around 1,500:1 (bare minimum for B2B SaaS), 2,000 emails gets you 1 qualified demo call per month. That's not consistent at all. For actual consistency - let's say you want 10 qualified calls per month - you'd need around 15,000 emails monthly. Most founders stick with 2,000 emails because they're worried about spam filters or think it's "enough to test." But you can't optimize with such little data: •⁠ ⁠2,000 emails gets you maybe 20-30 total replies all month •⁠ ⁠Takes months to know if your copy is working •⁠ ⁠No statistical significance on what's actually driving responses If you want to test cold email properly, you need minimum 50,000 emails per month. That gives you enough data points to see what's working within 4-6 weeks instead of 4-6 months. 1.⁠ ⁠Start with 3-5 domains, 2 inboxes per domain, send 25-30 emails per inbox daily. That gets you to 50,000 monthly volume with proper deliverability. 2.⁠ ⁠Track everything: open rates, reply rates, and most importantly - Email to Call Ratio. 3.⁠ ⁠Optimize based on the actual data The real issue isn't your messaging. It's just pure volume. You can't get consistent results from such low volume. If you want help building this infrastructure for your business, you can book a call here:

Actually a very useful idea Having built hundreds of agents id say this would be gold Certainly not a bad idea

Just post more content. Gurus say: Reality: Burnout Missed uploads Platform risk I built a workflow that clones a viral TikTok → rewrites it with AI → posts it to 9 platforms automatically Nobody does it better than me saving businesses since '05 haha

Small AI companies are outcompeting billion dollar comapnies This is an golden opportunity and the og @nvidia has given a blueprint on how to make an ai automation agency or atleast a business where you help implement ai agents Go implement a few !!!

Your AI gives you the same limp advice as everyone else’s. Here’s why: You tell it: “You are a marketing expert.” That’s a plastic title. It produces plastic thinking. I wrote a role definition that forces the machine to behave like a first-class strategist: 1) Copy this exact prompt 2) Paste into ChatGPT/Claude 3)Watch it deliver insights you won’t find in a handbook You are a ruthless strategist obsessed with results. You’ve spent 8 relentless years dissecting why some businesses erupt while others, selling the same wares, shrivel. You know the brutal truth: it is never the product. It is the psychological positioning—the way you frame it—that ignites trust, urgency, and desire in precisely the right people. As you analyze, you ask yourself: “What belief must I overturn for my audience to act?” You stop the moment you drift into clichés. You dig into the hidden barriers that paralyze action. Your obsession: uncover the single, blinding insight that makes every other tactic irrelevant. TASK: [Insert your specific request] CONTEXT: [Describe your situation, audience, goal] OUTPUT: Deliver the psychological breakthrough everyone else misses, then lay out a hard-edged, specific plan built on that insight. The difference is role definition. Generic role = generic advice. Specific role = Ogilvy-level patterns. Try it

most businesses use AI to “write more content” that’s why their results plateau the real wins come from using AI to design a content machine not just generate posts here’s the 3-part framework: The Insights Agent Prompt AI to act like your market’s top analyst and uncover fresh topics from industry forums, communities, and competitor channels Go beyond Google Trends find what your audience is already talking about The Research Agent feed the topics into an AI researcher It compiles structured briefs with stats, case studies, and recent developments so you start with context, not a blank page The Strategy Agent pass the briefs to another AI agent have it generate campaign plans, hooks, and format ideas tailored to each channel the result? a fully automated, insight-driven content engine where AI isn’t just writing for you. it’s thinking for you.

Stop calling yourself an “agency” if you’re just a freelancer with a fancy logo. Agencies don’t write emails at 2am. Agencies don’t fiddle in Zapier for 12 hours. Agencies sell deals, then make others run the machine. If you’re still the operator, you’re not the CEO. You don’t scale by working harder. You scale by disappearing from the work.

Claude VEO 3 NANO BANANA is absolutely INSANE This AI system turns VEO 3 into your personal ad factory producing unlimited UGC videos on autopilot And generates platform-ready creatives in minutes. Perfect for agencies & ecom operators who hate waiting on creators, editors, or expensive studios Instead of scripting, filming, and editing for weeks… Just give VEO 3 one simple prompt and get high-converting video ads at scale Each video comes with custom hooks, product demos, captions, and CTAs optimized for TikTok, Meta, and Instagram All from a single input Here’s what it does: → Creates unlimited UGC-style videos instantly → Writes hooks, CTAs & captions on demand → Auto-styles for TikTok, IG, and Meta formats → Generates 100s of variations in minutes → Uploads and organizes assets automatically → Runs 24/7 as your creative team replacement Built with VEO 3 claude AI Zero creators. Zero editors. Zero delays Want the complete system workflow? Comment “ADs” like this post and I’ll DM it over. (must be following)

VEO 3 Nano Banana n8n = Automated Ad Machine This system mass-produces TikTok, IG, and FB video ads using the newest AI video models. No actors. No editors. No bloated agencies. Just non-stop, UGC-style creatives designed to grab attention and convert. Perfect for ecom brands & agencies that live and die by constant creative testing. How it runs: → Upload your product catalog into Airtable → n8n pulls product data marketing angles → Nano Banana designs ad concepts layouts → VEO 3 turns them into high-quality video ads (hooks, demos, product-in-hand) → Each ad is auto-optimized for platform-specific virality → Airtable logs and tracks winners so you know what scales 24/7 ad production. Pennies per creative. And you keep 100% ownership. Want the complete workflow? Comment “VEO” Repost like this post, and I’ll DM it over. (must be following)

10 things I wish I knew before diving into AI automation (after building 30 workflows) Discussion Been deep in the automation game for the past year – here's what actually matters vs. what everyone talks about: 1. Start stupidly simple Your first automation should take 10 minutes, not 10 hours. I wasted weeks on complex builds when a simple "new email → Slack notification" would've taught me more. 2. Document your builds publicly Every automation you create is potential content. Screenshots, learnings, failures – it all becomes proof of expertise. I get more clients from sharing my process than from perfect demos. 3. Master the HTTP Request node first Seriously. Half the "limitations" people complain about disappear when you can build custom API calls. It's your Swiss Army knife for everything the built-in nodes can't handle. 4. Stop calling yourself an "automation expert" Everyone says that. Instead: "I help [specific industry] eliminate [specific pain point]." Specificity attracts premium clients who have that exact problem. 5. Your biggest wins come from saying no Turned down a $500 project last month because it wasn't aligned with my positioning. Client came back two weeks later with a $3K project that was perfect fit. Boundaries create value. 6. Error handling is where amateurs get exposed Everyone shows the happy path. Pros build for when APIs go down, data formats change, or users input garbage. Plan for chaos. 7. Share your failures, not just successes "Here's how I broke a client's workflow and what I learned" gets way more engagement than "Look at this perfect automation." Vulnerability builds trust. 8. The money is in ongoing optimization, not one-time builds Clients pay once for setup, monthly for "make it work better." Maintenance contracts beat project work every time. 9. Your network determines your net worth Other automators become referral sources, not competition. Help people in communities, share knowledge freely. Half my clients come from automator referrals now. 10. Build your own systems first Nothing proves automation expertise like having your own lead generation, content creation, and client onboarding automated. Practice what you preach.

right now, creators still piece things together prompts here, assets there, manual tweaks everywhere. But that’s already on borrowed time the next wave is prompt-to-production You’ll describe the concept in plain language The AI will: - Write the script - Generate scenes - Add voice, music, and effects - Edit, render, and publish without you touching a timeline the skill that wins won’t be “knowing the tools” It’ll be knowing what to make and why it matters text-to-image was the start text-to-video is here text-to-production is next your role is changing: stop thinking like an editor start thinking like a director

Open the left folder

Your next winning SaaS is hiding in your competitor’s 1-star reviews here’s how to use AI agents to find the gap: - Choose a top competitor on G2, Capterra, or Trustpilot - Deploy a scraper agent to collect their most recent 1- and 2-star reviews - Run an analysis agent to categorize complaints and rank them by frequency Prompt: “Group all complaints and show what % of reviews mention each issue” You’ll get outputs like: “38% complain about confusing UX” “24% mention bad onboarding” That’s your roadmap. Build the product they should’ve built, minus the flaws. no guessing no generic MVP Just engineering direct-response software based on pain points users are already screaming about.