Most buyers make offers based on a gut feeling about the multiple. That's how people end up owning businesses that pay them less than a job would. A simple Google Sheets model — five sections, maybe forty minutes of work — tells you exactly what you're buying before you commit a dollar.
Deal Alert AI is reader-supported. We earn commissions from affiliate links at no cost to you.
This post is based on a video from our Deal Alert AI YouTube channel. Watch the original or read the full breakdown below.
I have looked at thousands of online business listings. The single biggest difference between buyers who build wealth through acquisitions and buyers who end up trapped in a business that consumes their savings is not deal flow, negotiation skill, or even due diligence depth. It is whether they modeled the deal before they made the offer.
Not a complicated model. Not a leveraged buyout model with a debt waterfall and three tranches of preferred equity. A single Google Sheet with five sections that answers four questions: what does this actually return on my cash, can the business service the debt I'm taking on, what happens if revenue drops, and what is this worth when I sell it.
If you cannot answer those four questions with numbers, you are not investing. You are speculating with a spreadsheet-shaped hole where your analysis should be. This guide walks through the exact structure I use, the formulas that matter, and the thresholds that tell you to walk away.
Walk through any marketplace and you'll see listings priced at "3.2x SDE" or "42x monthly profit." Buyers anchor on that number instantly. A 2.8x deal feels cheap. A 4.5x deal feels expensive. That instinct is wrong often enough to be dangerous.
Here's a real comparison. Deal A: a content site at $400,000 with $125,000 SDE — a 3.2x multiple. Deal B: a SaaS product at $600,000 with $140,000 SDE — a 4.3x multiple. On multiple alone, Deal A wins. But the content site seller wants 90% cash at close and offers no financing. The SaaS seller will do 40% down with the balance on a five-year note at 7%. Suddenly Deal A requires $360,000 of your capital while Deal B requires $240,000. Deal A returns roughly $125,000 on $360,000 — about 34% cash-on-cash, though you've tied up almost all your liquidity. Deal B returns $140,000 minus roughly $85,000 in annual debt service, so $55,000 on $240,000 — about 23%, but you still have capital in reserve and the business is growing.
Which is better depends entirely on your capital position, your risk tolerance, and your time horizon. The multiple alone told you nothing useful. The model told you everything. That's the whole argument for spending forty minutes in Google Sheets before you send an offer. At Deal Alert AI we run this calculation on listings as they hit the market, precisely because the multiple headline hides more than it reveals.
Key insight: The multiple prices the business. The deal structure prices your return. Two deals at identical multiples can produce cash-on-cash returns 15 percentage points apart depending on seller financing terms, down payment, and amortization schedule.
We scan Empire Flippers, Flippa, Acquire.com and Quiet Light daily — scoring every listing. Start free.
The discipline that makes a model useful is separating inputs from calculations. Every number you might change for a different deal lives in one block at the top. Everything below it is formulas referencing that block. If you find yourself typing a number into a formula cell halfway down the sheet, you have broken the model and you will eventually make an expensive mistake.
Your input block needs: purchase price, trailing twelve month SDE, down payment percentage, seller note amount, seller note interest rate, seller note term in years, SBA loan amount, SBA rate, SBA term, your effective tax rate, and an assumed working capital reserve. That last one gets skipped constantly and it should not. Almost every online business needs three to six months of operating cushion post-close, and that cash is part of your invested equity whether you like it or not.
Use TTM SDE, not last year's, and not the seller's forward projection. If the seller hands you a P&L showing $180,000 SDE for calendar 2023 but the trailing twelve months through last month is $148,000, your input is $148,000. The business you are buying is the business that exists now. I'd go further — if there's a clear downward trend across the last three quarters, model the annualized run rate of the most recent two quarters instead. It's harsh. It's also accurate.
Color the input cells. Yellow, blue, whatever you like. When you open the sheet three weeks later to evaluate a new listing, you want zero ambiguity about which numbers are assumptions and which are outputs. This sounds trivial. It prevents the single most common modeling error, which is overwriting a formula and not noticing.
This section is short and it is the entire point of the exercise. Three lines matter.
First, total equity invested. That's your down payment plus closing costs plus working capital reserve. On a $500,000 purchase with 30% down, you're not investing $150,000 — you're investing $150,000 plus maybe $8,000 in legal and escrow plus $30,000 of working capital, so $188,000. Buyers routinely calculate returns on the down payment alone and overstate their return by 20% or more.
Second, annual debt service. Use the PMT function for each loan: =PMT(rate/12, term*12, -principal)*12 gives you annual payments. If you have both a seller note and an SBA loan, calculate each separately and sum them. Watch for interest-only periods or balloon structures — a seller note with two years interest-only followed by three years of amortization has wildly different cash flow in year one versus year three, and your model should show both.
Third, the two output metrics. Cash available after debt service equals SDE minus total annual debt service. Cash-on-cash return equals that number divided by total equity invested. DSCR equals SDE divided by total annual debt service. That's it. Everything else in the model exists to stress-test these two numbers.
Watch your DSCR floor. Any deal that models below 1.25x DSCR on base case assumptions is fragile. SBA lenders typically require 1.25x minimum for exactly this reason. If your model shows 1.15x, a 10% revenue decline puts you underwater on debt payments and you'll be funding the shortfall from personal savings while you try to fix the business. Do not talk yourself into it.
A single-scenario model is a wish. The sensitivity table is where you find out whether the deal survives contact with reality.
Build a 3x3 grid. Rows are SDE scenarios: bear case at SDE minus 10%, base case at flat SDE, bull case at SDE plus 10%. Columns are exit multiple scenarios: conservative 3x, market 4x, premium 5x. Each cell shows your total return under that combination. You can compute total return as either cash-on-cash in the operating years or as a rough IRR including the exit — I'd build both, but if you're keeping it simple, start with cash-on-cash in the bear row.
The bear-case row is the one that matters. If SDE drops 10% and your cash-on-cash return goes from 24% to 6%, that tells you the deal has almost no cushion — nearly all of your return is coming from the last slice of profit. If SDE drops 10% and you're still at 17%, you've bought a deal with structural margin for error. The gap between those two outcomes is usually the amount of leverage in the structure, which means it's negotiable.
Ten percent is not a pessimistic assumption for an online business. Content sites lose 10% of traffic from a single algorithm update. Ecommerce brands lose 10% when a competitor undercuts on Amazon. SaaS loses 10% when a large customer churns. Model a 25% decline too if the business has customer concentration above 20% or traffic concentration in one channel above 60%. Listings on Empire Flippers generally disclose traffic and revenue concentration in the listing itself, which makes it straightforward to pick a realistic downside input.
Cash flow is only half the return. The other half is the equity value you build by paying down debt and growing SDE, and you realize that half at exit.
Project business value at year three, year five, and year seven under each of your three multiple scenarios. Year three value at market multiple equals your projected year three SDE times 4x. Then subtract the remaining loan balances at that point. Google Sheets has no clean built-in for remaining balance, so either build a small amortization table or use =CUMPRINC() to sum principal paid through a given period and subtract from original principal.
Here's what surprises most first-time buyers: on a leveraged deal, the debt paydown is frequently a larger contributor to total return than SDE growth. Take a $500,000 purchase with $350,000 of debt across a seller note and an SBA loan. After five years of amortization you might have $190,000 remaining. That's $160,000 of equity created purely by making payments the business funded, before you've grown revenue by a dollar. Add flat SDE at a 4x exit multiple and your $188,000 of invested equity has become roughly $310,000 of exit proceeds plus five years of distributions.
Model the exit at a lower multiple than you paid. If you buy at 4.0x, model exit at 3.5x as your base case. Multiples compress. Buyer pools shift. The business is older at exit, and depending on the asset type, the market may value it less. Assuming multiple expansion in your base case is how buyers convince themselves that overpaying is fine. It isn't. If the deal works at a compressed exit multiple, you have a real margin of safety.
Key insight: On a properly structured leveraged deal, roughly half your total return comes from debt paydown funded by the business's own cash flow. This is why deal structure — down payment percentage, note terms, amortization — deserves as much negotiating energy as the headline price.
The last section is a single block at the top of your sheet — or on its own tab — that pulls every key metric into one view: purchase price, TTM SDE, multiple, total equity invested, annual debt service, cash available after debt service, cash-on-cash return, DSCR, and the year-five exit value range across your three multiple scenarios.
This exists so you can screen fast. When you're evaluating six listings in a week, you do not want to scroll through amortization tables to remember whether a deal cleared your threshold. You want to open the sheet, look at one block, and know. I keep a separate tab where each row is a deal and each column is one of these summary metrics — a simple deal comparison table that makes relative attractiveness obvious.
Add conditional formatting. Green if cash-on-cash is above 20% and DSCR is above 1.5x. Yellow between 15-20% and 1.25-1.5x. Red below those floors. It sounds like decoration. In practice it stops you from rationalizing a red deal into a yellow one during a phone call with a persuasive seller.
Work through these in order. The whole build takes forty to sixty minutes the first time and about ten minutes to adapt for each subsequent deal once you have the template.
Cash-on-cash return is the number that decides. Everything else is supporting evidence.
Below 15%, renegotiate or pass. That's not an arbitrary line. At 15% you're barely beating what a diversified index portfolio has returned historically, and you're taking on operational risk, concentration risk, and a job. An online business that returns 12% cash-on-cash while demanding fifteen hours a week from you is a bad trade — you'd make more working those hours somewhere else and putting the capital in an ETF. If the model says 12%, you have two options: negotiate a lower price, or negotiate better terms. Often terms are easier. Moving a seller from 20% seller financing to 40% seller financing can push a 12% deal to 19% without changing the price at all.
Above 30%, get suspicious. High modeled returns usually mean one of three things: the SDE is overstated, the price is below market because the business is declining, or you've missed a required expense. I have seen dozens of models showing 40%+ returns where the buyer forgot to add back an owner salary — the seller was working 30 hours a week for free, and once you hire a replacement at $45,000, the SDE drops and the return normalizes to 18%. Investigate before you celebrate. On Flippa, where the listing quality range is wide, an unusually attractive modeled return is more often a data problem than an opportunity.
The sweet spot is 18-28% cash-on-cash with a DSCR above 1.4x and a bear case that still clears 12%. That's a deal you can hold through a bad year without personal financial stress. That combination is rarer than you'd hope, which is exactly why fast screening matters — you need to look at many deals to find a few that qualify.
A model is only valuable if you use it at volume. Modeling one deal tells you whether that deal works. Modeling forty deals teaches you what a good deal looks like, what sellers will actually concede on terms, and where the market is mispricing certain asset types.
The practical bottleneck is input gathering. Pulling TTM SDE, understanding whether the seller will finance, and estimating working capital for each new listing takes longer than the modeling itself. This is where pre-screening earns its keep. We built Deal Alert AI to surface listings across the major marketplaces with the financial inputs already extracted and normalized, so you drop them into your model in under two minutes rather than digging through a prospectus for twenty. The acquisition model template covered in this guide is available free to subscribers.
Once you've modeled twenty or thirty deals, patterns emerge. You'll notice content sites at 3.5x with no seller financing consistently model worse than SaaS at 4.5x with 50% seller financing. You'll notice that Amazon FBA deals need a much larger working capital reserve than the listing implies, which crushes cash-on-cash. You'll notice which brokers' listings hold up under scrutiny and which don't. That pattern recognition is the real asset. The spreadsheet is just how you build it.
One final point. The model is not a decision-maker, it's a filter. It tells you which deals are worth spending forty hours of due diligence on and which to decline in ten minutes. Plenty of deals that model beautifully fall apart when you check traffic in Google Analytics or discover the top three customers are all the seller's former colleagues. But no deal that models badly gets better under due diligence. Start with the numbers, and use Deal Alert AI to make sure you're pointing them at enough deals to matter.
We scan Empire Flippers, Acquire, Flippa, and Quiet Light daily. The best sub-$500K businesses are gone within 48 hours.