Update, September 2026: the Deal Engine is no longer offered for install. It runs inside Serious Land Capital’s own deal review.
What I’m thinking about: The SLC Deal Engine was finished before its launch existed…and the one-day build that finally got it in front of you taught me more about marketing a premium product than the 2+ years of building it.
The engine was done, and I still couldn’t make myself build the launch.
Last week covered how the Deal Engine finally got reliable (the YouTube podcast version of this has gotten hundreds of views and climbing, best performing content so far…which matches what I’m most excited to present on…expect more of this).
That was the part I’d spent years on…but I still had to execute on the leads and conversions process: a sales page, an offer, payments, a demo video for the video sales letter (VSL), and a video edit of the kind I hadn’t touched in a decade. Huge pile of things I did not want to do (because of perceived drudgery).
And core to any offer is that people only care whether the thing or service works at a base level…nothing else matters if you can’t overcome that hump.
So the proof came first, before a word of copy. My manual review versus the engine, on the exact same deal, side by side in real time, on a property I’d never seen.
Anyone can demo a tool on a deal they’ve already solved (and build in hidden cheat sheets, or test basic conditions that won’t hold up under pressure). But coming in blind, with a manual read on a property in a purposely-chosen difficult, high-variance market, backed by a track record built on our own capital at title, is the receipt nobody can fake.
The equation the launch was built around
To organize the rest, I leaned on the framework I keep coming back to for any offer, Alex Hormozi’s value equation from $100M Offers (four levers…Dream Outcome and Perceived Likelihood of Achievement on top, Time Delay and Effort and Sacrifice on the bottom).
If you can remember your elementary math, the numerator you want as big as possible (in the case of the value equation) and the divisor as small as possible. Every objection I could steel-man, every proof point, every line of the sales page got sorted into one of those four buckets before anything got written.
Dream outcome
My dream for this business, for YEARS at this point, has been to keep removing myself from workflows until all that’s left is producing content, moving capital, and building/nurturing relationships.
The passive-income version everybody talks about (earning while you sleep, doing nothing) is generally not reality for the biggest financial goals IMO. But the operations piece, the part that demands your attention every single day, CAN be handed off, and most operators I know feel the same pull.
I don’t have to guess at that. When I described the target in a group chat with other land operators (plug in the parcel, zero additional input, get a pricing estimate and a diligence package out the other end), the replies came back the same way.
A framing I picked up recently from Varick Agents (an SF shop that builds custom AI agent systems for enterprise operations) has stuck with me…people don’t want another tool to do their work, they want their work DONE.
That’s the difference between the SLC Deal Engine and everything I built before it, and the standard I hold my own workflows to (and workflows I build for others).
Strip a land business down and the workflow is short. Leads come in, you review whether the deal is good enough, you fund it, you sell it. The review step in the middle is the most difficult, and I’d argue it’s the primary reason why so many operators have lost their shirts in this market or quietly exited (land valuations are complex, and for anyone who says otherwise, my counter is there wouldn’t be such significant market inefficiency and potential alpha if the process were basic…and there would be tons of spot-on land valuation tools available…there are not).
So the dream on offer is specific. Point the engine at a property, or at your qualified inbound lead flow, and come back to a sheet that says, “here’s the anticipated exit range, suggested purchase price is $X,” with extra DD items to dive into as needed (which the Deal Engine can drive almost the entirety of).
Perceived likelihood
A dream only counts for as much as the reader believes it’ll happen for them, which is where the race earns its keep.
If the engine matches my manual review, or outperforms it, on a deal neither of us has seen before, that’s about as strong as proof gets that it’s handling the hardest part of the job.
The receipts matter as much as the race, though. If someone with no track record ran the exact same exercise, the demo would prove nothing, because there’d be no reason to trust the manual read in the first place (let alone the engine that matched it).
That’s the piece most AI tool launches skip…they highlight the ‘Build’ and never show a ‘Check’ that cost anyone anything.
Time delay, effort, and sacrifice
Speed is king in every market I’ve operated in (it’s a major reason off-market land deals work, you offer speed of capital and the seller pays for it in price).
So the Deal Engine install is built to produce a real result the same day. I set it up on your machine personally (sub-10 min, as almost all of the install is now automated), we run your first deal together on the call, and by the time we hang up property reviews are off your team’s plate. (Self-serve installs are also available if preferred, but I’m inclined, and incentivized, to guide your first win with the Deal Engine as quickly as possible.)
The effort lever is the one I struggled with more than anything else.
Doing land underwriting properly, even with good tools, has meant endless back and forth…feeding listings in, supplying initial diligence, correcting the pricing mechanics mid-run. In the age of AI agents, customer expectations have moved past all of that, mine included.
The absolute minimum the engine needs from you now is a parcel ID, the state, and the county, and wiring it to your CRM so new leads process with zero input is a minor additional build.
The twin guarantees on the demo page are the same idea applied to risk.
Never to trick anybody, the point is making the offer a no-brainer for the operator it actually fits.
How the whole thing got built in about a day
With the equation sorted and the objections steel-manned, I handed the plan to Claude Code, and it came back with the VSL sequence (a ~2-3 minute intro, the full side-by-side Loom of the race, objection handling, then the guarantees and the call to action).
I recorded all of it, and then the part I’d been dreading showed up.
Ninety-ish minutes of footage, most of it me quietly clicking through comps next to a silent machine, and nobody is watching that. It needed speed-ups, voiceovers on the key moments, running timers on both sides of the split screen, branded title cards…the fine-detail editing that used to be its own week of work.
I described the problem to Claude Code (I have Riverside, iMovie, whatever is on this Mac…is there anything here that can do this?) and it pointed out that my Mac could run ffmpeg (a free, open-source command-line video tool, installed through Homebrew) and that it already had the full transcript. It could decide which passages to slow down because they explained the Deal Engine, which to speed through because they were dead air, and cut the whole thing to length on its own.
First cut took ~90 minutes of footage down to ~20.
I sent back the nits (a couple of spots clipped my voice, the timers and captions faded into the frame) and asked it to pull our brand system from Claude Design and rebuild the cards, captions, and timers to match. One more edit session, and…done. It built the website to the same brand direction too, Calendly wired in, Stripe connected, click-by-click install instructions included.
Ten years ago I built video sales letters by hand, which took a couple of weeks, and the quality was far worse.
Even a year ago, hiring a professional editor for the speed-ups and branded cards alone would’ve run a few days and at least ~$1K.
Now the public side of a premium launch fits inside a day, nasty little details (time-lapse editing chief among them) included.
And apparently I’m not the only one who stumbled into this.
Since recording, I’ve learned there’s an open-source project called video-use (from the team behind Browser Use) that packages the exact workflow…drop raw footage in a folder, tell Claude Code what you want, get a finished file back. What I improvised in an afternoon is already becoming a category (and actually I took it upon myself to improve the ‘video-use’ tool by forking my own version which utilizes free transcript-processing tools instead of relying on costly API credits from ElevenLabs…and you can get my version here, or download the skill file which accomplishes the same thing).
None of the above existed in usable form until recently, and to be clear, we’re operating on the frontier here…pushing the latest model releases to their absolute limit, inside Claude Code, to make this Deal Engine product work at all (I’ve never been more excited to keep showing up every day, knowingly building something the world has never seen…and it’s actually working).
What still had to come from me
The value equation script couldn’t be delegated.
What determines value for the user, which objection is real and which is noise…that requires human attention from someone who’s lived the workflow. Being my own number one customer made it easier, because I only had to build something for myself first (and my own expectations are nothing short of exacting).
Early response has been encouraging. Two customers are fully in (our internal KPI was three installs by the end of Nov. just to hit bare-minimum financial targets), and marketing has barely started.
This week, I’m presenting the Deal Engine live in front of hundreds of operators (preparing for that has been consuming almost every ounce of attention).
Maybe I eat my words over time and this lands smaller than the early feedback (and my own excitement) suggests.
Nobody controls outcomes in a market, only the effort and the learning that go into each week, and either way 2+ years of iterations finally cracked the original dream…and without hyperbole I’d put my skills as an AI workflow engineer against anyone at this point, having never upskilled myself this fast in my life so far, and progress keeps accelerating.
Even if it only solves my own dream, that’s fine, as this has been the highest and best use of my time recently. I don’t think it stops there, though…most people dream the same thing. The difference has always been who acts on what’s required.
If you’re ready to become the kind of land operator whose underwriting (and business) runs without you, watch me race my own engine on a live deal and judge the output for yourself.
I set the engine up on your machine personally (or you self-serve the install using my guide), and we run through your first deal together…if it doesn’t look like the video, you don’t pay.
