Buying a $2M business with 10% down
and growing it with AI (i show everything)

In the last thirty days, for just one of our clients, we sent out ninety applications to buy businesses. Twenty-seven brokers wrote back. We signed six NDAs, and three confidential books landed in his inbox — while he was still doing his day job.

And I'm gonna show you the whole thing today. The actual board we run, the real documents, the real numbers. Including the deals we told him to walk away from and why.

And I know this because over the last eighteen years we've bought, invested in, operated or supported hundreds of business transactions — more than a billion dollars' worth. And I used these same strategies to sell Rollups.com last year to Naval Ravikant's company.

By the end of this you're gonna know how to buy a two million dollar business with about two hundred grand — either yours, or raised from us or from other investors we can connect you with. Where the deals actually come from. How to tell in about twenty minutes whether a business is a good deal or not. And then what AI we install in the first ninety days after you own it, which is honestly the half nobody talks about.

One thing before we start. Everything I show you today, I'll give you. Just comment the word ASSETS below and I'll send it over. There's nothing behind it.

So this is it. Everything on the left is how we find and buy a business. Everything on the right is what we do to it after you own it.

And I'm gonna go through every single box. Left to right. No skipping.

Box one. And honestly this is the one everybody skips, which is exactly why they waste a year.

Before you look at a single listing, you write down what you'll actually buy. That's your buy box. Cash flow floor, price range, geography, three to five industries. That's it.

And the cash flow floor matters more than people think. Below about two hundred and fifty grand, the owner is doing everything himself, and you're not buying a business — you're buying yourself a job with debt attached.

Three to five industries, not one. Because one industry might throw up two listings a month and you'll starve waiting.

And this is what it turns into. A real buyer profile.

Geography. What they can actually finance. Their background and operating experience — because a guy who spent thirty years building something does care who ends up with it. Their criteria. And who's standing behind them.

This is the document a broker reads before deciding whether you're worth a phone call. Which is why we write it properly instead of making you fill in a form.

Next one's boring but it matters. Accounts.

We open and run accounts for you on every platform. That's twenty-plus logins — marketplaces, broker portals, auction sites.

Why does that matter? Because a broker looking at a brand new account with no history treats that completely differently than an established buyer. It's the difference between getting the book and getting ignored.

Okay so now deal flow. And there are exactly three ways into a business. Most people only ever try the first one, and then they complain everything good is gone.

Way one is listed. Everybody knows BizBuySell. But look at this — that's sixteen marketplaces we're scanning every single day. BizQuest, DealStream, BusinessBroker, LoopNet, Axial, Sunbelt, Transworld, Murphy. Not one.

And here's the thing about listed deals nobody tells you. It's not about finding something nobody's seen. It's about being first.

A broker reads the first three enquiries and basically stops reading properly after ten. On a good listing that window is under forty-eight hours.

So the AI agent we built scores every listing against your buy box, and applies for you automatically. Up to fifty a day. Before you'd have even opened your laptop.

And this is what comes back. That's a real client's board.

Colour coded — under LOI, under contract, sold, passed with the reason written down. So nothing gets lost and you can see the whole thing at a glance.

And for every single deal we're pulling location, revenue, EBITDA, cash flow, year established, employee count, seller type, who the broker is. All of it, before anyone spends time reading anything.

Way two. Pre-market. And this is the one that actually matters.

Pre-market means the broker's got the business, he hasn't listed it yet, and he sends it to buyers he already knows. That's where the deals with no competition are.

Right here — twenty-seven hundred brokers in our database. Twelve hundred with a direct line. Heaviest in Florida, California and Texas.

And it took us ten years to build those relationships. That's not something you can go and buy.

So what we run is five emails over two weeks. And it's a specific order.

Introduction. Then credentials and your buy box. Then something genuinely useful about their market — so now you're a peer and not a lead. Then on day ten, the direct ask. And then day fourteen, should I close your file.

And I'll just show you the actual emails, because this is the bit people get wrong.

Email one — who the buyer is, what he's funded for, proof of funds available, signs NDAs same day. Short.

Email two — the buy box written out. Cash flow, price, states, type, must-haves. So the broker knows exactly what to send.

Email three is the one everybody skips. No ask at all. Just something useful about his market — we've looked at eleven HVAC deals this quarter, nine were retirement sales, seven had no manager under the owner, and that's killing what buyers will pay. That's it. No ask.

Then day ten. The direct ask. Anything pre-market? I'll sign an NDA today and get you proof of funds within the hour.

And that's the email that produces deals. But it only works because emails one to three already made you a known, funded, easy buyer.

And here's a real reply. Broker comes back — actually yes, I've got a roofing and construction business going to market in about three weeks, three hundred and eighty-five grand of earnings, owner retiring, GM staying on. Send the NDA and I'll get you the package before it goes live.

That's a pre-market deal. Nobody else is looking at it.

And this is the campaign actually running. Sent from your address, on warmed domains, capped at thirty to forty a day so your inbox stays healthy.

Look at the reply rate on day ten. Twenty-one replies, seventeen of them positive. That's the highest of any step — because by then you're not a cold email any more.

And this is the folder those go into. Memoranda, executive summaries, P and Ls, income statements, balance sheets, financial recasts.

Look at the range — a restaurant group doing six point seven million, a fifty-year flooring business, a general contractor, a med spa, a diagnostic lab. That's what deal flow looks like when it's actually working.

Way three. Off-market. Straight to owners who never listed at all.

And I'll be straight with you — this is the hardest one, and that's exactly why almost nobody does it.

Look at what's actually involved. You scrape the owner list from public data. You enrich it with verified contact details through Apollo and Clay. You buy ten to twenty separate sending domains. You warm those domains for two, three weeks so your mail lands in an inbox instead of spam. Then you send, and you monitor it every day.

Five separate systems that all have to work together. Most people quit at step three.

But that's where the deals with zero competition are. The guy who's sixty-six, real customers, real staff, real profit, and he's never once spoken to a broker.

And these are the databases we're pulling from. SourceCo, two hundred million businesses. Grata, sixteen million companies. Inven, twenty-three million. Kumo, SourceScrub, PrivSource.

This is private company data that will never appear on a listing site.

And then the outreach itself. Three weeks, five emails, straight to the owner. Opening with a qualified buyer, closing with a reason to reply either way.

A hundred to a hundred and fifty owners a week.

Now, everything from here goes out under your name, from your email address. Applications, NDAs, all of it.

Because the second it looks like an agency in the middle, you drop down the pile. The broker wants to talk to the buyer.

NDAs signed and returned the same day, every time. Speed is the whole game at this stage.

Right. This is the most important part of the whole video. If you skipped ahead, come back here.

A memorandum arrives. It's forty pages. And you have to understand what that document actually is — a broker paid somebody to write it to make that business look as good as it possibly can. It's a sales document. It is not a neutral report.

So we run the same eight questions on every single one. Price and what's actually included, because inventory and real estate are often quoted separately. Three-year average earnings — including the bad year, not just the good ones. What it does in one sentence. Top five risks. How involved is the owner. Is there a manager. Six questions specific to this deal. And number eight — who are you gonna have to hire the day after you close.

Number eight is the one that kills people.

And here's what that looks like on a real deal. Look at these notes.

Earnings fell from eight hundred and forty-one grand to five hundred and fourteen. Labour running at thirty-five percent when the ceiling in that industry is about twenty-five. A hundred and nineteen thousand dollars of expenses that can't be categorised. And the owner stepped back last year, which is why the numbers dropped.

None of that's in the headline. All of it's in the financials if you actually read them.

And this is the report we produce. Sixteen pages, thirteen sections.

Executive summary. Revenue quality and trends. Add-back audit and normalised earnings. Margin and cost structure. Working capital. Financial verification checklist. Operations. Customers. Competitive position. Transition terms. Growth plan. Risk register. And a recommendation with a proposed structure.

That's the document that stops you buying something that looked fine.

And then a short version too, because nobody reads sixteen pages on every deal. Three pages. Headline numbers, financing structure, the three-year trend, top risks, owner involvement, the questions to ask, and who you'll need to hire.

That's the one that saves you the most time.

Let me give you the five things that kill a deal. If you take nothing else from this video, take these.

One. Trailing twelve months is negative. Doesn't matter how good the years before look — you're buying a decline, and the bank sees it too. Pass.

Two. Add-backs that won't survive a lender. Owner salary, fine. One-off legal fees, fine. Officer supplies, family payroll, vague personal expenses — the lender throws those straight back in. And every dollar they reject comes off the earnings your price was based on.

Three, and this is the big one. The hiring haircut. Owner works forty-five hours a week, no manager. That's a hundred and fifty grand a year. Take it off before you work out the multiple. Most deals that look like three times are really four and a half.

Four. Customer concentration. If the top five customers are more than about thirty percent of revenue, one phone call can halve that business. Above fifty percent you're not buying a company, you're buying a relationship — and the relationship is with the guy who's leaving.

Five. Debt service coverage under one point five. Which is the next box.

That whole checklist — comment ASSETS below and I'll send it to you.

So now the maths. And really there's only one number that decides whether a deal works.

Debt service coverage ratio. Take the annual earnings after you've taken out a manager's salary. Divide it by everything you owe each year across every loan. That's your ratio.

Under one two five, a lender says no. And if they don't, you should. One two five to one five is tight — only works if that revenue's genuinely stable. One five to two is the target, that's where you've got room for a bad quarter or a truck breaking. Above two, comfortable — go and check you haven't missed something.

And here's the model itself. Yellow cells on the left are what we change — price, earnings, loan percentage, interest, years, seller note, equity, working capital. Right side calculates.

And the black bar at the bottom is the recommendation. With a suggested call that a human can override, because sometimes the numbers work and the deal still stinks.

And this is the bit everybody gets wrong. Everyone assumes you need two million dollars to buy a two million dollar business. You don't.

You put in ten percent. The seller carries ten percent as a note — and that's good for you beyond the money, because a seller with a note still cares how the handover goes. The bank does the other eighty on an SBA seven-A. And that ceiling just went up to ten million.

And here's the sentence I want you to remember. A bank will not fund your startup. It will absolutely fund you to buy a business that already makes money.

Why? Because thirty years of cash flow is collateral, and an idea isn't.

So while everybody's grinding trying to build something from zero and can't get a dollar out of anyone, there's a guy two towns over with real customers, real staff, and nobody to hand it to. And a bank will give you eighty percent of him.

Also, two things nobody budgets for — working capital, which is roughly sixty to eighty percent of your equity injection again, and twenty-five to fifty grand of legal and accounting at closing. Neither is in the asking price.

Now capital. Because before a broker sends you a single financial statement, they want proof of funds.

And if you haven't got one ready, you're just out. Three other buyers do, and they get the book first.

So we generate it per deal. Right here.

And a proper one has seven things. The buyer's full legal name matching the NDA exactly. A specific dollar figure, not 'sufficient funds'. The stated purpose. A funding timeline in days. A verification contact they can actually ring. An expiry date. And real letterhead with a real signature.

And this is what actually lands in his inbox. Referenced, dated, pre-qualified to a number, valid ninety days, with someone to call and check.

And then this one. This is a real bank pre-qualification.

That's an actual lender who has reviewed the buyer's personal financial statement, his liquidity, his credit profile and his management experience — and put a number in writing before he's even made an offer on anything.

You walk into a conversation with that, and you're not a tyre-kicker any more. You're a buyer.

And if you haven't got the ten percent, or the deal's bigger than you can cover — this is the room we open.

Twenty-four hundred active acquisition investors. Fifteen hundred with verified contact details, a hundred and fifty-five with a direct dial. So a deal that needs equity has somewhere to go.

And thirty-five lending firms who do independent-sponsor and first-time-buyer debt. Sixty-one named contacts. Typically two million EBITDA and up.

That's for when the deal's too big for SBA, which happens more than you'd think once you get going.

Then we structure the offer.

Every deal side by side — price, earnings, multiple, coverage ratio, annual cash flow. And a go or pass on each one, with the letter of intent and the full report one click away.

And we send you the no's as well. You should see what we protected you from.

And when it's a yes, the letter of intent gets written. Price and terms. Seven risk contingencies — key person, owner transition, non-compete, supplier concentration, add-back reconciliation, inventory verification, employee classification. Assets in and out. Diligence and closing timeline. Exclusivity.

Four pages, and it's ready to sign.

And then the introduction goes out. From your email address, with your buyer profile and your calendar link attached.

So the broker turns up to that call already knowing you're serious. You're not selling yourself on the phone — that's already done.

And this is what you actually see every day.

Brokers contacted. Replies received. NDAs signed. Proof of funds letters sent. Memoranda in. Deals analysed. Follow-ups sent. Calls booked for you.

Then underneath, every actual reply — who wrote in, what deal, what they sent. And every document attached.

Every single day. Including the quiet ones. When nothing happened, the report says nothing happened. That's the point.

And a portal you can open any time. Every deal we've vetted, why we like it, the numbers, the broker's details, and a four-step checklist of what you need to do next.

Which is usually just: make the call.

Okay. So that's how you buy one.

And most firms stop right there. They get you to the closing table, shake your hand, and you drive to a building full of people you've never met, with a phone that rings all day and a business that lives inside one guy's head.

That's where people actually lose money. Not in the purchase. In the first ninety days.

So let me show you the other half.

Right. Part two. And I'm gonna go left to right again, in the order a customer actually moves through your business.

First one, getting found.

So we build the site. Fast, works properly on a phone, financing visible, service-area pages for every town you cover.

And crucially it books. Somebody lands at eleven at night on a Sunday, they pick a time and it writes straight into your calendar. Most contractor sites have a contact form that emails an inbox nobody opens.

Then Google. Paid at the top with the Guaranteed badge, and number one in the map pack underneath. Both, not one.

And the map is the one that compounds, because those calls are free.

And this one's new, and almost nobody local has thought about it yet.

When someone asks ChatGPT who to call for a roof repair in their town — you want to be the name that comes back. And right now that window is wide open, because your competitors don't know it exists.

Box two, lead generation. And this is where we actually create demand.

And here's how the ads get made, because this is the part owners dread.

We write the script. You either film it yourself in about ten minutes, or we use a human-looking AI presenter — your call. We edit it, cut it for each platform, and send it to you. You approve it or you ask for a change. Then it goes live.

Your total time is about ten minutes a month approving things.

And we don't report cost per lead. Cost per lead is a number agencies use to look good.

We report cost per booked, completed, paid job. Those are different numbers and only one of them is real.

And then organic. One video gets made, and it goes out everywhere — Instagram, TikTok, YouTube, Facebook, LinkedIn, X. Every day.

And that's what it looks like after ninety days. One point four million views.

And the second payoff nobody expects — it fixes hiring. In a trade where finding technicians is the hard bit, being the visible local company means applicants come to you instead of you chasing them.

Box three. And this is where most of the money is leaking right now.

One in five calls in these businesses goes unanswered. Eighty-five percent of those people never ring back. And forty-one percent of calls come in after hours.

So here's a real one. Quarter to nine on a Sunday night.

Answered in two rings. Qualified. Address confirmed. Booked for Monday morning between eight and ten. Confirmation text sent before the call ended.

That job was four thousand one hundred and eighty dollars. And before this was installed, that call went to voicemail.

And everything lands in one place. Calls, texts, the website form, Facebook, Instagram, Google Business, email. One thread per customer with the full history.

So nothing sits unanswered just because it came in on the wrong app.

And it books against real technician availability and your service area. Grouped by geography, so your guys aren't driving across the county twice a day.

And you watch it happen. Every call, every booking, every escalation, as it lands. Then a summary at close of business.

Average speed to lead — four point three seconds.

Box four. Follow-up. And this is where thirty to fifty grand a year quietly dies in most of these businesses.

Tech quotes eight grand for a job. Customer says let me think about it. And then nobody ever calls back.

So every unsold quote goes into a sequence. Day zero, day three, day seven, day fourteen, day thirty. With financing options in there, because a lot of the time it's not a no, it's a cash flow problem.

Look at that conversation — 'bit more than I expected honestly'. And then financing at a hundred and eighty a month, and it's a yes.

Close rate went from forty-four percent to sixty-one.

And then reminders, so they're actually there when your guy turns up.

Confirmation on booking with a reply-to-confirm. Two days out. The night before — anything we should know, which catches problems early. Morning of, with the tech's name and photo. And then twenty minutes out from the tech himself.

No-shows went from eighteen percent to six. And a no-show is a paid technician driving to nothing, which is the most expensive hour in the business.

Box five. Sales. Because getting them booked is only half of it — somebody's got to close the job.

So every sales call gets recorded and transcribed. And the AI scores it.

Look at this one. He presented three options, asked about budget early, offered financing — all good. But he never actually asked for the sale. And his talk time was sixty-eight percent, which is too high.

And then the coach sends him a note that evening. 'Good work on the three options. Next time, once she says let's do that one — stop talking. You kept selling for another forty seconds.'

That's coaching from what actually happened, instead of from what somebody remembers.

And the pricing itself becomes a system. Flat-rate book, so it's not different depending on who turns up. Three options on every quote — and the middle one gets picked about fifty-four percent of the time, which lifts your average ticket on its own. Financing offered above a threshold.

Plus a scorecard per technician. Close rate, average ticket, callback rate, review score. So you can see who's closing and who needs help. That's usually worth more than hiring anyone new.

Box six. The two cheapest sources of work in any service business, and almost nobody systemises either one.

After every completed job — how did we do, one to five.

Four or five goes straight to Google with a one-tap link. Anything lower comes to the owner privately instead, while you can still fix it. And then the referral ask, with fifty dollars each way.

Look at that thread. Five stars, straight onto Google, then 'my neighbour was asking, sent it to her'. That's two jobs from one text.

And this is ninety days of that. Four point one to four point nine. A hundred and eighty-seven new reviews.

Which is exactly why you're now first in the map. It compounds.

And then the list you already own. Fourteen hundred past customers, worked on a seasonal calendar.

Ninety-six thousand dollars from that last campaign. Zero ad spend. These are people who already bought from you once.

Box seven. And this is the one owners tell me changes how the job feels.

One screen for today — calls, jobs booked, revenue, average ticket, and where every one of them came from. Live.

And then the month closes itself. Revenue by line, gross margin, operating costs, EBITDA, cash position. Plus it flags the service lines quietly running below margin, which is usually at least one.

No waiting six weeks for the bookkeeper to tell you how you did.

Box eight. And this one's optional, but it's the one people get excited about.

All of this sits behind one screen. And if you want, you can stop clicking through it and just ask.

How did we do this week. What's still unpaid. Book the Thursday slot. It answers, and it does it.

And then the last box. Which is really the whole point.

Look at this curve. A small owner-run business trades at about three times its profit. Mid-size with real management in it, around six. Inside a large group, eight to ten or more.

Same work. Same trade. The only thing that changed is size.

So yes, the systems make you more money. But they also move you along that curve — and the move along the curve is worth more than the extra profit. Most owners never see the right-hand side of it, because they never get big enough on their own.

Which is the other thing we do. Run the system for ninety days. If the numbers are there, we make you an offer — we invest, your company joins our group, and you take cash out now while keeping a stake that's valued at the group's multiple instead of yours.

And if it's not a fit, no problem. The system's yours to keep.

So that's the whole thing. Buy box, three ways to find deals, the analysis, the numbers, the capital, the offer — and then everything we install to grow it.

Two things and I'll let you go.

One. Everything in this video, I'll give you. The eight-question checklist. The five deal killers. The sixteen marketplaces. The funding structure. The proof of funds template. The broker email sequence. Comment ASSETS below and I'll send it over. No email, no upsell.

Two. If you'd rather not build all of this yourself, that's what we do. Link's in the description. Thirty minutes, we look at what you've got and where you want to get to, and we'll tell you honestly if it's not a fit. We say that a lot, actually.

Either way — go and buy something real. Thanks for watching.