Most buyers fall in love with a listing before they've done ten minutes of real math. The good ones do the opposite — they try to kill the deal fast, and only the survivors get an offer. Here's the exact six-step sequence I use on every listing.
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.
By Sophal Lanh, Founder of Deal Alert AI
I look at somewhere between 40 and 80 online business listings a week. Most of them get about ninety seconds of my attention. That's not laziness — it's discipline. The single biggest mistake I see new buyers make is spending three weeks in diligence on a business they should have rejected in the first two minutes, then feeling emotionally obligated to buy it because of all the time they sunk in.
The framework below is a funnel. Each step is designed to kill the deal. If the deal survives all six, you submit a Letter of Intent. If it dies at step two, you saved yourself thirty hours. The whole point is to make the cheap checks first and the expensive checks last.
I'll walk through each step with the actual numbers and thresholds I use. This isn't theory — these are the filters running behind Deal Alert AI every morning when new listings hit Empire Flippers and Flippa.
Before you open a single spreadsheet, you look at four things: business type, revenue trend direction, asking price relative to monthly profit, and age of the business. That's it. Four data points, all of which are visible on the public listing page without signing an NDA.
Business type tells you what kind of operator you'd need to become. A content site monetized by display ads and affiliate links is a fundamentally different job than a Shopify store with inventory and supplier relationships, which is different again from a B2B SaaS with a support queue. There's no universally "best" type — there's the type that matches your skills, your capital, and how many hours a week you actually have. If you're a developer with a day job, a physical products business with 3PL headaches is a bad fit no matter how attractive the multiple looks.
Revenue trend direction is the one that kills the most deals. Sellers list businesses at the peak, not the trough. If you're looking at a trailing twelve-month chart and the last three months are lower than the three before them, you're not buying a business — you're buying someone else's decline. Sometimes there's a legitimate seasonal explanation. Usually there isn't. My rule: three consecutive months of declining revenue is an automatic reject unless the seller can show me a specific, verifiable, already-reversed cause.
We scan Empire Flippers, Flippa, Acquire.com and Quiet Light daily — scoring every listing. Start free.
Beyond the four data points, there are three specific things that make me close the tab immediately. These aren't negotiating points. They're exits.
Declining revenue for three or more consecutive months. Covered above, but worth repeating because buyers rationalize this constantly. "It's just a Google update, it'll recover." Maybe. But you're paying today's price for a recovery that may never come, and the seller is exiting precisely because they don't believe it will. If the person who knows the business best is leaving, ask yourself what they know that you don't.
Business younger than twelve months. A business needs to survive at least one full annual cycle before you can trust its numbers. You need to see the Q4 spike and the January hangover. You need to see how it handled at least one algorithm update or one ad platform policy change. Under twelve months, you're not buying a business — you're buying a promising experiment at a business valuation. Under twenty-four months, be skeptical. Over thirty-six months, you're looking at something with actual durability.
Owner claims zero hours per week. This is the one that sounds like a feature and is actually the loudest possible alarm. Nothing runs at zero hours. Somebody is answering emails, somebody is approving content, somebody is dealing with the supplier who shipped the wrong SKU. If the owner says zero, one of two things is true: they're not counting the work they do (which means you'll discover it at 11pm on your second week), or there's a team member doing it who may not stay after the sale. Ask for the actual weekly task list. If they can't produce one in under 24 hours, that tells you how the business is really run.
SDE stands for Seller's Discretionary Earnings. It's net profit plus the owner's salary plus one-time expenses plus any personal expenses run through the business. It is not revenue, and the multiple you see quoted is almost always a multiple of SDE, not revenue. Confusing these two is the fastest way to overpay by 4x.
The math is simple: asking price divided by annual SDE. A business asking $240,000 with $6,000 in monthly SDE is $240,000 ÷ $72,000 = 3.33x. That's a reasonable multiple for a stable content or affiliate business. A business asking $240,000 with $6,000 in monthly revenue is a completely different — and probably insane — proposition.
The harder work is verifying the add-backs. An add-back is an expense the seller argues you won't have to pay. Some are legitimate: a one-time website redesign, legal fees for a trademark filing, the seller's health insurance. Some are aggressive nonsense: "conference travel" that was actually a family vacation, or a $2,000/month contractor add-back where the seller claims you'll do the work yourself. Here's the test — if you'd have to pay for it to keep the business running at the same level, it is not an add-back. Strip out every add-back you don't believe, recalculate SDE, and recalculate the multiple. I've seen a listing go from an advertised 3.1x to a real 4.4x after removing three bad add-backs. That's the difference between a good deal and a bad one.
One more thing: check the SDE trend, not just the SDE total. A business with $72,000 trailing-twelve-month SDE where the last quarter annualizes to $50,000 is not a $72,000 SDE business. Weight recent months more heavily. I typically build three numbers — TTM SDE, last-6-months annualized, and last-3-months annualized — and value off the lowest of the three unless there's a clear seasonal reason not to.
If you're financing with an SBA 7(a) loan — and a lot of buyers in the $200K to $2M range are — the lender will run a Debt Service Coverage Ratio calculation. You should run it first, because if the deal doesn't clear DSCR, the bank won't fund it and the whole conversation is academic.
DSCR is your annual cash flow available for debt service divided by your annual debt payments. Lenders generally want 1.25x minimum. Some want 1.5x. Here's a real example. Say the business does $10,000/month SDE, or $120,000/year. You're buying at $400,000 with 10% down, so you're financing $360,000 over 10 years at roughly 11.5%. That's about $5,060 a month in debt service, or $60,700/year. But you also need to pay yourself — call it a $60,000 salary. Now your cash available for debt service is $120,000 minus $60,000 = $60,000, and your DSCR is 60,000 ÷ 60,700 = 0.99x. Dead on arrival.
Change one variable and the picture flips. Same business at a $320,000 purchase price: financing $288,000, debt service around $4,050/month or $48,600/year, DSCR of 1.23x. Still tight. Take a $48,000 salary instead of $60,000 and you're at $72,000 available, DSCR 1.48x — fundable. This is why running DSCR early is so valuable: it tells you your actual maximum offer price, not the price the listing says. You walk into negotiation with a number derived from math instead of vibes.
Screenshots are worthless. PDF exports are worthless. You need read-only access to the actual Google Analytics 4 property and Google Search Console, logged in from your own machine, poking around yourself. Any seller who refuses this after an LOI is telling you something.
Once you're in, four things matter. First, the organic versus paid split. A business that's 90% organic search traffic has a durable but algorithm-fragile moat. A business that's 90% paid has a cost structure that can be destroyed by a CPC increase but is at least controllable. A business that's 90% one social platform is a business with a landlord who can evict you. Know which one you're buying.
Second, look for anomalies. Sudden traffic spikes with no corresponding revenue spike are a classic bot-traffic tell. Traffic that dropped 30% on a specific date and never recovered probably lines up with a Google core update — go look up the date. Traffic that's suspiciously smooth, with no daily variance at all, is often not human. Real traffic is noisy.
Third, page-level concentration. Export the top 50 pages by sessions. If one page drives 60% of traffic and that page targets a single keyword, you own a keyword, not a business. I like to see the top page under 20% of total traffic and the top ten pages under 55%. Fourth, check the referring domains in Search Console. A backlink profile made entirely of paid guest posts from the same three networks is a link penalty waiting to happen.
Same principle as traffic: get into the source system. Stripe, PayPal, Shopify, Amazon Seller Central, the affiliate network dashboard — whatever actually processes the money. Screen-share sessions are acceptable if the seller won't grant access, but you drive the mouse and you pick which months to open.
Go month by month for the last 24 months and compare each figure against what the seller put in the P&L. You are looking for two things: months where the numbers don't match, and refund or chargeback activity that was netted out quietly. On a Shopify store, pull the refund rate specifically — a store showing 4% refunds in the P&L and 11% in the actual data has a product quality problem you're about to inherit.
If there's recurring revenue, churn is the number that matters more than MRR. A SaaS doing $15,000 MRR with 3% monthly logo churn is a very different asset than one doing $15,000 MRR with 9% churn. At 9%, you replace your entire customer base every eleven months, which means you're not buying a subscription business — you're buying a sales machine that happens to bill monthly. Pull cohort retention if the platform supports it. Ask for gross and net revenue retention separately.
Finally, check revenue concentration. One affiliate partner at 70% of income, one wholesale client at half of revenue, one Amazon SKU carrying everything — these are single points of failure that should either reduce the multiple you're willing to pay or end the conversation entirely. I discount concentrated revenue heavily: if the top source is over 50%, I mentally knock 0.5x off the multiple I'd otherwise offer.
Here's the whole framework as a working checklist. Run it in order. Stop at the first hard fail.
If all six steps pass cleanly, submit the LOI. Don't wait. Good deals at fair multiples get multiple offers within days on both Empire Flippers and Flippa, and the buyer who moved on day one with a clean, well-reasoned LOI usually wins over the buyer who spent three weeks being thorough. Being fast and being rigorous are not opposites — that's exactly what the framework is for.
If one or two things came back soft — a slightly aggressive add-back, a 6% churn rate, a page-concentration problem — that's a renegotiation, not a rejection. Quantify the issue in dollars and present it that way. "Your SDE includes $18,000 of contractor add-backs that I'll have to keep paying, so real SDE is $54,000, not $72,000. At the 3.3x you're asking, that's $178,000, not $240,000." Sellers respond to arithmetic far better than they respond to opinion. About a third of the time you get a meaningful price move.
If something hit a hard red flag, walk, and walk without agonizing. There will be another deal next week. There is always another deal next week. The buyers who blow up are almost never the ones who missed a good opportunity — they're the ones who talked themselves into a bad one because they'd already invested emotional energy. Sunk cost is the most expensive line item in this business.
The bottleneck for most buyers isn't the analysis — it's finding enough deals worth analyzing in the first place. That's why I built Deal Alert AI: it scans new Empire Flippers and Flippa listings every morning, scores them against these exact filters, and sends you only the ones that survive. You get the first-pass filter done before you open your laptop, so your time goes to steps three through six on deals that actually deserve it. Start with the framework, run it consistently, and let Deal Alert AI handle the volume.
We scan Empire Flippers, Acquire, Flippa, and Quiet Light daily. The best sub-$500K businesses are gone within 48 hours.