There's a survey out from Indeed.. 80,000 workers across eight countries.. and the finding that stuck with me is this: most of the time people save with AI gets redirected into doing more of the same work. More of the same. At the same price. Or lower. That's the trap I'm watching funnel builders walk into right now. AI shows up, cuts your build time in half, and the instinct is to pass that savings along. Charge less. Move faster. Stay competitive. It feels like the logical response to a world where anyone can spin up a funnel in an afternoon. But think about what's actually sitting inside that recovered time. Uptime monitoring for every client funnel. Weekly dead link scans. Page speed checks. SEO health reports. Automated social content pipelines. Lead nurturing sequences. App maintenance. The kind of ongoing infrastructure work that a full-service agency charges a $2,500–$4,000/month retainer to deliver. A bigger agency doesn't just build a funnel and hand it off. They run it. They watch it. They catch the broken order button before the client emails you in a panic on a Saturday. They produce a monthly report that makes the client feel taken care of.. and keeps the contract renewing. You couldn't do all of that alone before. The bandwidth wasn't there. It's there now. So before you send your next proposal.. think about what you can stack in that offer that you couldn't have touched six months ago. Monitoring. Reporting. Automation. Ongoing maintenance. Services that justify a number your old hourly rate never could. The time AI gives back isn't a discount to pass along. It's the capacity to build a bigger offer and charge accordingly. 🚀 - James

Funnel Builders.. Every client who built their own app with AI is eventually going to hand it to you. They built it in Lovable, or Replit, or whatever tool promised them a working product in an afternoon. It looked finished. It launched. People used it. And then something broke. And now it's your problem.. They can't tell you how it was built. The AI generated it.. there was no architect, no decision trail, no documentation. The client doesn't know what's under the hood. They just know it's not working. So you're looking at something that was summoned, not constructed. And you have no foundation to stand on when you try to read it. Now you're stuck.. with this knowledge gap.. one nobody warned you was coming. The funnel builder world taught you to build inside safe, visual editors. That was enough. Until it wasn't. And now apps are landing on your desk that assume a foundational understanding of how software actually works.. build configurations, authentication flows, integration handoffs, deployment pipelines.. that you were never given and never needed before now. What changes everything is building enough of that foundation to be able to triage. To look at a broken app and identify the category of the problem. Is this a domain issue? An integration issue? A code issue? A "this needs to be rebuilt" issue? Those are four completely different conversations, and knowing which one you're in changes everything about what happens next. Funnel builders who can do that triage are going to be the ones worth calling. The ones who can take the handoff, read what's in front of them, and know which direction to move.. or know exactly who to bring in when the direction is clear. That foundation is buildable. And it's closer than you think. 🚀 - James

I've been thinking about experts recently.. people putting out courses and communities.. So many are worried about AI taking their marketshare.. For some, it's happening already. An LLM has a lot of general purpose data in it and.. the really cool thing.. is that chat experience it has.. The person can ask a question, get an answer, ask a follow up question.. get another answer. Admit they're confused or unsure.. get suggestions for how to get unstuck. It's a rich, dynamic, two-way conversation that.. if we're being honest.. is a lot nicer than a course. Thinking about a course.. it's one-way communication. The expert does their best, using story, metaphor, everything they have, to convey the necessary information to achieve transformation. But.. if we're being honest.. most experts are not great educators. Being good at a thing doesn't mean you are good at getting other people good at that thing.. That's its own skill. So, faced with a course, or the subscription they already have.. I can see why they pick AI over you. So.. bummer situation right? Are experts done? Is it all AI from here on out? I don't think so. Here's what I've been thinking about.. and playing with. The one thing that AI doesn't have a drop of is lived experience. We know they hallucinate.. they make up things that sound good but.. just ain't right. But even if they stay straight.. it's a conglomeration of every possible solution to the problem, right? We've learned that niching down is super important, so you can attract people that will relate to you.. values wise.. culturally.. so they'll be more likely to get the same win you got. If AI is duct-taping a "mega-result" framework from everyone else's stuff.. then it's like.. the opposite of niching.. And, it's a unproven theory. So.. instead of just letting AI go nuts with its own theories, we constrain it. Instead of AI replacing the Expert, we put AI into the Expert's Framework. What does this look like? Well. It doesn't look like a "GPT" of all of your books and SOPs.

A lot of app ideas never leave someone's head. The resistance used to be valid because you had to know how to program, and even if you did, you had to spend time doing it, or pay someone to do it for you. Now with AI, you can spin up a minimum viable product in minutes. That will make you more excited than you've ever been about the idea. I was talking to a friend recently who had a solid idea.. combining AI with a student tracking system for an after-hours school he runs. Gamification, progress tracking, push notifications. Real thing. Worth building. But he'd never built an app before. And that blank page before anything exists.. it's its own kind of stuck. So I just started talking. Walked him through what a product designer would actually ask. Platform, users, data, security, how it makes money, what the MVP looks like versus what comes later. And somewhere in that conversation, a plan showed up. That's when I realized.. this is a prompt. Whether you're the expert with the idea, or you're the funnel builder who is the hands behind someone else's vision, the starting point is the same. You need a plan before you build. And the fastest way to get one is to let an AI act as your product designer and interview you into it. So I built a starter prompt that does exactly that. You fill in three blanks: *My app helps [type of person] solve [problem] with [your rough idea].* That's it. From there the LLM takes over. It asks you one question at a time.. covering platform, user accounts, security, data storage, notifications, monetization, anything a product designer would want to know. Then it produces a PRD with a phased approach. Phase 1 is your MVP. Phase 2, 3, and beyond is everything that comes after. The whole thing comes out of a conversation. You don't need to know the technical vocabulary. You just need to answer questions. I'm dropping the full prompt in the comments. 🚀 - James

This one is for my fellow agency owners and funnel builders.. You built your reputation on knowing how this stuff works. Not theory. Actual knowing. The kind that comes from being inside a hundred builds, understanding why a funnel breaks at step three, knowing which button nobody checks on mobile. That was the identity. Technician. The person who makes it work. Then agents showed up. And they're better at the technical execution than most of us ever were. I'll include myself in that. Faster, tireless, no learning curve on the repetitive stuff. If your whole value proposition is "I can build this".. that proposition just got very complicated. Here's the thing though. The agent doesn't know what done looks like. It doesn't know to check every button on desktop and mobile. It doesn't know to verify that a purchased product gates the right step in the funnel. It doesn't know the thousand foundational things you've absorbed across years of real builds.. the failure modes, the edge cases, the details that look fine until a client launches to 10,000 people and something quietly breaks. That knowledge didn't expire. It's actually worth more now. Because the person who can direct an agent, catch what it misses, build apps that go beyond what any platform restricts you to, wire everything back to a proven system that has worked for over a decade.. that person is operating at a level they never could have before. The technician identity was always underselling the real asset. The real asset is judgment. The years of pattern recognition. The ability to look at something and know instantly whether it's actually done or just looks done. Agents are execution. You're the standard. So use them. Build faster. Build apps that solve problems no template can touch. Monitor the health of funnels like infrastructure, not one-off projects. Make sure everything is fast, indexed, and showing up the way it needs to in a world where AI is deciding what's worth surfacing. The ceiling on what you can deliver just came off.