Revenue gets people excited. SDE gets people paid. If you can't rebuild a listing's SDE from bank statements and Stripe exports in under an hour, you're not doing due diligence — you're trusting a stranger's spreadsheet.
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.
Deal Alert AI is reader-supported. We earn commissions from affiliate links at no cost to you.
By Sophal Lanh, Founder of Deal Alert AI
Every week I look at dozens of listings across the major marketplaces. The listings that get the most clicks always lead with revenue. "$42,000/month site for sale." That headline sells. It also tells you almost nothing about whether the business is worth buying, because revenue is what customers pay in — not what you keep.
The number that actually determines your price, your loan terms, your payback period, and whether you can quit your job in three years is SDE — Seller Discretionary Earnings. This article breaks down what SDE is, how to compute it correctly, which add-backs are legitimate, how to verify the number from primary documents, and how to turn SDE into a defensible offer price.
Seller Discretionary Earnings is the total financial benefit a single full-time owner-operator extracts from the business in a year. That includes the net profit shown on the P&L, the salary the owner pays themselves, and any personal or one-time expenses run through the company that a new owner would not have to repeat. In plain English: SDE is the money the business puts in your pocket if you run it yourself.
This is different from EBITDA, which is the standard metric in larger transactions. EBITDA assumes the business pays a market-rate manager and reports profit after that cost. That works when you're buying a $20M company with a management team. It doesn't work for a $400K content site where the "management team" is one person working fifteen hours a week from a laptop. For small online businesses — roughly anything under $5M in enterprise value — SDE is the industry standard, and it's what every serious broker will present.
Revenue lies because it hides cost structure. Two businesses can both do $42,000/month in revenue. One is a SaaS product with 88% gross margins and $6,000/month in total costs, producing roughly $36,000 in monthly SDE. The other is a dropshipping store spending $28,000/month on paid ads, $9,000 on cost of goods, and $2,500 on a virtual assistant team, producing about $2,500 in SDE. Same top line. One is a $1.2M business; the other might be worth $75,000 on a good day. If you anchor on revenue, you will systematically overpay for low-margin businesses and miss high-margin ones.
We scan Empire Flippers, Flippa, Acquire.com and Quiet Light daily — scoring every listing. Start free.
The formula itself is simple: Net Profit + Owner Salary + Legitimate Add-Backs = SDE. The difficulty is never the arithmetic. It's deciding what belongs in each bucket.
Start with a clean example. A productized service business reports $8,000/month in net profit after all operating expenses. The owner also pays herself a $3,000/month salary that is recorded as an operating expense on the P&L. Because a new owner would step into that same role and capture that same salary, it gets added back. SDE is $11,000/month, or $132,000 annually. If the market multiple for that business type is 3.0x annual SDE, the asking price lands near $396,000 — not the $288,000 you'd calculate off net profit alone, and not some fantasy number based on revenue.
Now add a layer of realism. Suppose the same business also spent $9,000 one time on a website rebuild, pays $400/month for the owner's personal health insurance through the company, and expensed a $4,200 conference trip that was half business and half vacation. The rebuild is a genuine one-time cost and gets added back in full. The health insurance is a personal expense a new owner wouldn't inherit — add it back. The conference trip is a judgment call; a reasonable buyer adds back half. Annual SDE becomes $132,000 + $9,000 + $4,800 + $2,100 = $147,900. At 3.0x, that's a $443,700 valuation — a $47,000 swing driven entirely by add-back decisions.
That swing is why add-backs are the single most negotiated part of any online business deal. Every dollar of add-back you accept costs you three or four dollars in purchase price. Every dollar you successfully challenge saves you the same. At Deal Alert AI we treat the add-back schedule as the real contract negotiation, not the LOI.
A legitimate add-back meets one test: the expense would not exist for a new owner running the business normally. Owner salary and payroll taxes on that salary qualify, because the buyer replaces the owner's labor with their own. Personal expenses run through the company — the owner's car lease, personal phone, home internet, family health premiums, a gym membership coded as "wellness" — qualify, though they should be modest. True one-time costs qualify: a logo redesign, a legal settlement, a website migration, the cost of a failed product launch that won't be repeated. Interest on seller-owned debt qualifies, since the debt doesn't transfer. Non-cash items like depreciation and amortization qualify, because they aren't real cash leaving the business.
Then there are the add-backs sellers try to sneak in. The most common is marketing spend framed as "growth investment." A seller will say, "I spent $60,000 on Facebook ads last year to grow — you don't need to spend that." If the revenue those ads produced is included in the SDE, you cannot remove the cost. That's double counting, and it's the single most frequent inflation tactic I see on marketplace listings. The same logic applies to content production for content sites: if the site's traffic depends on publishing 20 articles a month, the writer cost is an operating expense, not an add-back.
Other red-zone add-backs include contractor costs the seller claims they'll "stop needing" post-sale, software subscriptions that are actually load-bearing for operations, recurring "one-time" expenses that appear in three consecutive years, and family members on payroll who genuinely perform work. If the seller's mother handles customer support 20 hours a week for $1,500/month, that $18,000 is not an add-back — it's a cost you will inherit or have to replace at market rate, which is probably higher.
Never accept a seller's P&L as fact. A P&L is a claim; bank statements are evidence. Your job in due diligence is to rebuild the SDE number from primary sources and see whether it matches within a few percent. If your rebuild comes in more than 5% below the listing, either you're missing information or the listing is inflated — and you need to know which before you wire money.
Start with revenue. Pull 24 months of raw payment processor exports — Stripe, PayPal, Shopify Payments, Amazon settlement reports, or ad network statements for content sites. Sum the net deposits after processor fees, refunds, and chargebacks. Then match those totals against actual bank deposits month by month. Any deposit in the bank that doesn't appear in the processor data needs an explanation; any processor revenue that never hit the bank does too. For advertising-based businesses, log into Mediavine, AdThrive, or Ezoic directly during a screen share and compare the dashboard to the claimed figures.
Then verify expenses, which is where most inflation hides. Get the business bank statements and credit card statements for the same 24 months and categorize every recurring charge. You'll find subscriptions the P&L omitted, contractor payments that weren't disclosed, and "one-time" costs that show up quarterly. Finally, request tax returns — Schedule C for a sole proprietorship, Form 1120S for an S-corp. Tax returns are filed under penalty of perjury and sellers systematically minimize income on them, so if the tax return shows more profit than the P&L, that's usually honest. If the P&L shows dramatically more than the return, ask why in writing.
Here's the exact sequence I run on any deal before making an offer. It takes two to four hours for a well-documented business and flushes out most inflation problems before you spend money on legal or a formal audit.
Buyers who run all ten steps rarely get burned. Buyers who skip to step ten because the listing looked clean are the ones writing painful forum posts eighteen months later. Our scanning tools at Deal Alert AI automate the front half of this list, but steps five through ten still require a human brain and a phone call with the seller.
The first red flag is an owner claiming near-zero hours. "I work two hours a week" appears on thousands of listings and is true on perhaps five percent of them. Ask the seller to walk you through their last full week, task by task, with timestamps. Ask who handles a refund request, a Google algorithm update, a supplier delay, a chargeback dispute. If they can't answer specifically, either they work more than they claim or the business is running on autopilot toward a cliff.
The second is inconsistent monthly patterns without explanation. Legitimate businesses have seasonality, and seasonality has a story — Q4 for ecommerce, January for fitness, back-to-school for education products. What should worry you is a flat 24-month record with one or two enormous spike months, or revenue that grows in a suspiciously smooth straight line. Real revenue is lumpy. Fabricated revenue tends to be tidy.
The third cluster: add-backs above 30% of SDE, "one-time" costs appearing in multiple years, a single customer or single traffic source representing more than 30% of the business, recent changes to the ad network or payment processor that make historical comparison impossible, and a seller who resists screen shares. That last one is nearly disqualifying. An honest seller with clean books will happily log in and show you the dashboards. A seller who insists on emailing you PDFs and won't do a live walkthrough is telling you something important.
Once you have a verified SDE figure, price follows from a multiple. In the online business market, multiples are typically quoted as either a multiple of annual SDE or a multiple of monthly SDE — brokers like Empire Flippers quote monthly, where a "40x" listing means 40 times monthly SDE, or roughly 3.3x annual. Get the units right before you compare anything, because a 3x annual business and a 30x monthly business are the same business.
As a rough map of the current market: content and affiliate sites generally trade around 30–40x monthly SDE, ecommerce and Amazon FBA around 30–45x depending on brand strength and inventory, SaaS with genuine recurring revenue and low churn at 40–70x, and services or agency businesses at the low end, 20–32x, because they carry high owner dependency. Multiples move with age, revenue concentration, traffic diversity, growth trend, and how transferable operations are. A five-year-old site with diversified traffic and documented SOPs earns a premium; an eighteen-month-old site with 80% of traffic from one Google query does not.
Run the return math before you anchor on any multiple. At 3.0x annual SDE, you recover your capital in three years assuming performance holds flat — an implied 33% annual return. At 4.5x, payback stretches past four years, which only makes sense if you have high conviction in growth. Then subtract a realistic operator cost. If the business produces $150,000 in SDE but genuinely requires 25 hours a week, and hiring that out costs $50,000, your true passive yield on a $450,000 purchase is $100,000 — 22%, not 33%. Both numbers can justify a deal; only one of them is honest.
Not all marketplaces verify financials to the same standard, and knowing the difference saves you enormous amounts of wasted diligence. Empire Flippers runs its own vetting process before a listing goes live, including P&L reconstruction and traffic verification, which is why their listings carry higher multiples and a smaller inventory. You're paying a premium for pre-screened deal flow, and for buyers doing their first or second acquisition, that premium is usually worth it.
Flippa operates as an open marketplace with far more volume and far more variance. Some listings have verified Google Analytics and connected Stripe accounts; others are self-reported with nothing behind them. The upside is real: because vetting is lighter, well-run businesses sometimes list at multiples two to eight points below what the same asset would fetch through a full-service broker. The tradeoff is that you have to do the verification work yourself, which is exactly why the checklist above exists.
The practical problem for most buyers isn't analysis — it's coverage. Good deals with clean SDE and a fair multiple get taken within days, often by buyers who saw the listing the hour it went live. Manually refreshing marketplace pages isn't a strategy. That's the specific problem Deal Alert AI was built to solve: we scan the major marketplaces daily, parse the stated financials, compute the implied multiple against category benchmarks, and flag listings where the numbers deserve a second look. You still run your own diligence — nobody should ever outsource that — but you start from a filtered list instead of a firehose.
The discipline underneath all of this is simple. Compute SDE yourself from primary documents. Build your own add-back schedule. Price off the number you verified, not the number you were shown. Do that consistently and you'll pass on nine deals out of ten, which is exactly the point — the tenth one is where the money is.
By Sophal Lanh, Founder of Deal Alert AI
We scan Empire Flippers, Acquire, Flippa, and Quiet Light daily. The best sub-$500K businesses are gone within 48 hours.