Most SaaS founders think the software is the moat.
Spend two years building the perfect tool. Add features. Out-engineer the next guy. Pray nobody copies you.
I used to believe that too.
Then I watched a video from Brett (The Brett Way) that put words to something I'd been feeling for a while: the tool isn't where the money is anymore. The outcome is.
He opens with a number that stuck with me. One of his creators got on the phone, made his third cold call of the day, and walked away with $850 from a barbershop. For a website. That the creator built in 2 minutes with AI.
Two minutes of work. $850 in the bank. And the barber was happy.
Stupid simple, right? That's exactly the point.
Services are the new software
Here's the part that should make every SaaS founder sit up.
Sequoia (yeah, that Sequoia) published a piece arguing that for every $1 spent on a software tool, roughly $6 gets spent on the service that delivers the exact same outcome.
Read that again.
$1 on the software. $6 on the human (or the agency, or the freelancer) doing the work with it.
So you've got two ways to play the AI wave:
1. Sell the software. Which means you're competing with the next model update, the next YC-backed clone, the next free tool that does 80% of what you do.
2. Sell the outcome. The finished website. The booked calls. The cleaned-up data. And every time the AI gets better, your margins go up, not down.
That second one is the unfair game. When you sell outcomes, every Claude or GPT release is a tailwind. When you sell software, half of them are a threat.
I'm not saying torch your SaaS. I'm saying the line between "software company" and "service company" is blurrier than it's ever been, and the founders who notice early get a window.
The arbitrage window (and why it closes)
Brett calls it opportunity arbitrage. I like that.
It's when a new technology changes how work gets done, but the market hasn't repriced it yet. For a short stretch, you can use the new tool to create a wildly disproportionate amount of value.
What used to take weeks now takes minutes. What used to need a five-person team now needs one person and a good prompt.
So you deliver faster. You serve more clients. Your margin goes up. That's the arbitrage.
But here's the catch with any arbitrage: it doesn't last.
The people who move early eat. The people who wait until it's "proven" show up to a crowded table. Same thing happened with Facebook ads in 2013, with dropshipping in 2016, with newsletters in 2020. Early is everything.
Right now, in 2026, a kid with a Claude subscription can fulfill an entire service before the prospect even says yes. That's the window. It's open today.
The playbook Brett actually ran
What I respect about the video is he doesn't just theorize. He builds the whole machine on camera with the Claude desktop app and the Chrome extension. Here's the shape of it.
Step 1: Prospecting on autopilot. He told Claude his goal (sell websites to local restaurants in Kansas City), had it write a skill file with a simple scoring system (bad website or no website = good lead, skip the big chains), and set a scheduled task to pull 25 new leads into a spreadsheet every morning at 7am. Name, Instagram handle, phone, email. Done while he sleeps.
Step 2: Build the deliverable before they ask. Another skill file scrapes each restaurant's menu, photos, and branding, then generates a custom website. His hack here is gold: grab a brand-guidelines template off Canva, export it as a PDF, hand it to Claude, and the site suddenly looks like a real designer made it. Hosted free on Vercel, so every lead gets a shareable link. Scheduled for 8am daily.
Step 3: Outreach that runs without you. The Chrome extension takes over Instagram, finds each lead, sends the pre-built website link plus a follow-up with a calendar booking link. 9am, every day. He wakes up to booked sales calls.
Step 4: Get paid in 60 seconds. He spins up a Whop checkout link for $1,000, sends it after the call, and the client pays by card, bank transfer, or installments. No subscription, no setup hell.
Notice the staggering: 7am leads, 8am websites, 9am outreach. He gives each step an hour because the agents need time to run. That detail is the difference between "cool demo" and "actually works."
If you're a founder who's allergic to building lead-gen machinery, this is the kind of thing I'd point you to before you hire anyone. I wrote more about the manual version of this in how to get your first 100 SaaS customers, but the new playbook is mostly setting up the right business automations and letting them run.
What this means if you sell SaaS
You might be reading this thinking "cool, but I'm not starting a website agency."
Fair. But sit with the lesson, not the tactic.
The lesson is: people pay for outcomes, not access. Your prospects don't want a dashboard. They want the thing the dashboard produces.
So a few moves for SaaS founders:
→ Add a done-for-you tier. If your tool does X, sell "we'll do X for you" at 5x the price. That $6 of every $7 is sitting right there. Some of your best customers will gladly pay it.
→ Use the service to fund the software. Charge for outcomes now, learn exactly what people want, then bake it into the product. This is the reverse of what most people do, and it's safer. I broke down the full version of this in how to turn an agency into a B2B SaaS.
→ Stop competing on features. If the next model update can clone your feature, it was never the moat. The relationship, the outcome, the trust, that's the moat.
I paid $13,000 for a mastermind once when I was doing $15-20K/month. Stupid decision on paper. Six months later I was at $75K/month. I didn't pay for the Slack group. I paid for the outcome. People do this every single day.
The catch nobody puts in the thumbnail
I'd be lying if I sold you the "AI does everything, you do nothing" dream. Brett is honest about this too, which is why I trust him.
The number one skill in the AI world is taste. The AI builds the website in two minutes, sure. But it can't tell you if it's good. That's still you.
Three things that'll save you:
1. Check the real data. Brett's first auto-generated site only pulled 5 menu items off the actual menu. AI is very good at confidently making stuff up. Cross-check your first few deliverables by hand before anything goes out the door.
2. Hold a high bar. If you deliver garbage fast, you've just built a machine for producing garbage faster. Care about the work. That's what turns one $1,000 client into three referrals.
3. Don't spam. Automated Instagram DMs at scale will get your account nuked. The platforms' rules are murky and they change. Move deliberately, not desperately.
And the pricing? Brett says $500 to $3,000 for a single website right now, with retainers and upsells (SEO, Google reviews, online ordering) waiting on the other side. The website is just the foot in the door. The relationship is the business.
FAQ
Is "services are the new software" actually a real thesis or just a YouTube hook?
It's a real argument. Sequoia published a piece making the case that the most defensible AI businesses might not be the ones selling the software tool, but the ones using it to deliver the outcome. Their framing: for every $1 spent on a tool, about $6 is spent on the service delivering the same result.
Do I need to be technical to run the AI services playbook?
No. The whole point of the Claude desktop app plus the Chrome extension is that a non-technical person can set up skill files and scheduled tasks in plain English. The skill that matters isn't coding, it's taste and follow-through.
How much can you charge for an AI-built website?
In the video, $500 to $3,000 for a one-off site, depending on the client and the polish. One creator closed a barbershop for $850 on his third cold call. The bigger money is in retainers and upsells after you've earned trust.
Should a SaaS founder do this instead of building software?
Not instead. Alongside. Add a done-for-you tier, sell the outcome at a premium, and use the cash and the customer conversations to make your actual product sharper. Services can fund and inform the software.
Won't this opportunity get saturated?
Eventually, yes. That's how arbitrage works. The early movers capture most of the value before the market reprices it. The window is open now, which is exactly why it's worth moving on this quarter, not next year.
One more thing
The reason I keep coming back to stories like this is simple. The founders winning right now aren't the ones with the most features. They're the ones who figured out what people actually pay for and delivered it faster than everyone else.
That's the whole game.
If you want more of these breakdowns (real founders, real numbers, what's actually working in 2026), the Profitable Founder Podcast is where I dig into it every week with bootstrapped founders doing $100K to $10M a year.
Go listen. Then go build something people will pay for the outcome of.
