Every SaaS listing you read will lead with ARR. Almost none of them will show you the four components of MRR that actually determine whether that ARR is durable or evaporating. Here's how to read subscription revenue like an operator instead of a spectator.
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.
If you have spent any time browsing SaaS listings, you have seen the same headline structure a hundred times: "$480K ARR SaaS, 92% gross margin, 3% monthly churn, asking $1.6M." Three numbers and a price. Most buyers read that, do a quick division to get the multiple, and move on to the next listing.
That is a mistake. ARR is a summary statistic. It compresses twelve months of customer behavior into one number and throws away everything that tells you whether the business is compounding or quietly bleeding out. Two SaaS businesses can both report $480K ARR and be worth wildly different amounts — one might be worth $1.9M, the other might be worth $700K, and the difference lives entirely in the MRR components that never make it into the listing headline.
This guide walks through what ARR and MRR actually measure, how they differ, how to break MRR into its four moving parts, how to verify every number during due diligence, and when to value a SaaS business on ARR versus SDE. If you are evaluating subscription businesses, this is the foundation everything else sits on.
Monthly Recurring Revenue is the sum of all predictable, contracted subscription revenue a business collects in a given month. If a SaaS product has 500 customers each paying $100 per month, MRR is $50,000. Annual Recurring Revenue is that same figure annualized — $600,000. That is the entire mathematical relationship. ARR = MRR × 12. Nothing more complicated than that.
What trips people up is the word "recurring." Both metrics deliberately exclude anything that does not repeat on a predictable schedule. One-time setup fees are excluded. Implementation and onboarding charges are excluded. Custom development work, migration services, training sessions, consulting retainers that are not on subscription terms — all excluded. A business that collects $50,000 per month in subscriptions plus $12,000 per month in setup fees does not have $744K ARR. It has $600K ARR and a services line that happens to generate $144K annually.
This distinction matters enormously in acquisitions because services revenue is not worth the same multiple as subscription revenue. Subscription revenue renews without effort. Services revenue requires you to go win it again every single month, and it typically carries a fraction of the margin. Sellers who blend the two into a single "ARR" figure are inflating the metric that drives their asking price. I have seen listings where 30% of the stated ARR was non-recurring, which meant the real multiple the buyer was paying was 40% higher than advertised.
We scan Empire Flippers, Flippa, Acquire.com and Quiet Light daily — scoring every listing. Start free.
If you have bought content sites, ecommerce stores, or newsletters, you are used to SDE multiples — seller's discretionary earnings times some number between 2.5x and 4.5x. SaaS breaks that convention. The market prices most software businesses as a multiple of revenue, specifically ARR, and the multiples look absurd by comparison until you understand why.
A stable, modestly growing SaaS business with reasonable retention typically transacts somewhere in the 3x to 5x ARR range in the sub-$5M market. A genuinely fast-growing SaaS business — 40%+ year-over-year growth with net revenue retention above 100% — can command 5x to 8x ARR. Compare that to a content site at 3.5x SDE and the numbers look insane, until you remember that a 90% gross margin SaaS business converts revenue to profit at a rate a content site never will, and that subscription revenue renews automatically while affiliate revenue depends on Google's mood next Tuesday.
The revenue-multiple convention also exists because SaaS earnings are deliberately suppressed during growth phases. A founder who is reinvesting every dollar into paid acquisition, engineering hires, and content might show near-zero SDE on $700K ARR. Valuing that on SDE would produce a nonsense number. Valuing on ARR captures the underlying asset — a book of recurring contracts with high margins — rather than the accounting output of one particular reinvestment strategy. Marketplaces like Empire Flippers list SaaS deals with both figures for exactly this reason, and the gap between them tells you a lot about how the seller has been running the business.
Here is where most buyers stop doing work and where the real analysis begins. Total MRR is a net figure produced by four independent forces pushing in opposite directions. If you only look at the net, you cannot tell whether a flat MRR line means a stable business or a leaky bucket being refilled by expensive paid acquisition.
New MRR is revenue from customers who did not exist last month. Expansion MRR is additional revenue from existing customers — upgrades, seat additions, usage overages, plan changes upward. Churned MRR is revenue lost from customers who cancelled entirely. Contraction MRR is revenue lost from customers who stayed but downgraded. Net new MRR = New + Expansion − Churned − Contraction.
Now run the scenario. Business A adds $8,000 in new MRR each month, $1,000 in expansion, and loses $2,000 to churn and $500 to contraction. Net new MRR is $6,500. Business B adds $3,000 in new MRR, $5,000 in expansion, and loses $1,500 to churn and $500 to contraction. Net new MRR is $6,000. Both look nearly identical on a growth chart. But Business B's existing customers are generating $5,000 in expansion against $2,000 in losses — the customer base grows on its own. Business A depends entirely on filling the top of the funnel, which means its growth is only as good as its next ad budget.
Never accept a seller's MRR spreadsheet at face value. Spreadsheets are opinions. Payment processor data is fact. The single most important step in SaaS due diligence is getting screen-share or read-only access to the actual billing system and pulling the numbers yourself.
In practice this means Stripe, Paddle, Chargebee, or a Baremetrics/ProfitWell dashboard sitting on top of one of them. Ask for a 24-month MRR export broken down by component — new, expansion, churned, contraction — plus a customer-level subscription table showing signup date, plan, current status, and lifetime value. Twenty-four months matters because twelve months hides seasonality and lets a seller time the sale right after an unusually good acquisition quarter.
Then reconcile. The Stripe MRR line for each month should tie to the subscription revenue line in the P&L, allowing only for processing fees, refunds, failed payment recovery timing, and any revenue collected outside the processor. If the reconciliation is off by more than a few percent in any month, you need a specific, documented explanation. "The bookkeeper categorized it differently" is not an explanation. It is a signal to slow down.
The ARR-versus-SDE question comes up on nearly every SaaS deal, and the honest answer is that it depends on the margin and growth profile. There is no universal rule, but there is a reliable framework.
Use ARR multiples when gross margins are 70% or higher and the business is growing at a meaningful clip — say 20%+ year over year. In that profile, the recurring revenue book is the asset, market comparables exist, and the multiple ranges are well established. A $600K ARR business at 85% gross margin growing 35% annually with 2% monthly churn is a textbook ARR-multiple deal, and you should benchmark it against similar listings rather than against SDE-based businesses.
Use SDE multiples when the business is flat or declining, when gross margins have compressed below roughly 60% because of heavy infrastructure or support costs, or when a large chunk of revenue is services rather than subscription. A no-growth SaaS product with $400K ARR, 55% margins, and $150K SDE is functionally a stable cash-flow business, not a growth asset. Pricing it at 4x ARR would be $1.6M for $150K of earnings — a 10.6x SDE multiple that no rational buyer should pay. The correct frame is 3x to 4x SDE, or roughly $450K to $600K.
My preference is to run both calculations on every deal and pay attention to the gap. When the ARR-derived value and the SDE-derived value are within 25% of each other, the pricing is probably sane. When ARR-based value is triple the SDE-based value, the seller is selling you a growth story, and you need to independently verify that the growth is real, durable, and not purchased at a customer acquisition cost that never pays back. This dual-lens approach is exactly how Deal Alert AI scores SaaS listings — every SaaS deal that hits the platform gets evaluated on both frameworks so you can see the gap immediately instead of computing it by hand across dozens of listings.
Before you send an LOI on any subscription business, work through this list in order. Each item takes between ten minutes and an hour, and collectively they eliminate the vast majority of bad SaaS deals before you have spent real money on professional diligence.
Metrics are only useful if they change what you are willing to pay. Once you have clean component-level MRR data, you can build a defensible valuation instead of arguing about multiples in the abstract.
Start with adjusted ARR — subscription revenue only, verified against the processor. Apply a base multiple appropriate to the growth rate: roughly 3x for flat, 4x for 15–25% growth, 5x to 6x for 30%+ growth with clean retention. Then adjust. Negative net churn earns a premium of half a turn to a full turn. Customer concentration above 25% in the top five costs you half a turn. Monthly churn above 5% costs a full turn or more. Heavy founder involvement in support and sales costs you a turn because you are buying a job, not an asset. Meaningful platform dependency — everything running through one app store or one integration partner — costs at least half a turn.
Then run the SDE cross-check. Take trailing twelve month earnings, add back the owner's salary and genuinely discretionary spend, and see what a 3x to 4x SDE multiple produces. If your ARR-based number is $1.4M and your SDE-based number is $500K, do not simply average them. Understand why the gap exists. Sometimes it is legitimate reinvestment into growth that you would continue. Sometimes it is a business that spends nearly every dollar of gross profit to stand still.
Finally, get comparables. Look at what actually sold, not what is listed. Browse closed SaaS transactions across marketplaces — Flippa for the smaller end of the market and Empire Flippers for vetted deals above $100K — and note the relationship between stated ARR, retention quality, and final sale price. Asking prices tell you what sellers hope for. Sale prices tell you what the market pays.
ARR and MRR are not interchangeable buzzwords. MRR is the operating metric — the thing you manage week to week, broken into four components you can individually improve. ARR is the valuation metric — the thing brokers put in headlines and the thing multiples get applied to. Buyers who only look at ARR are reading the summary of a book they never opened.
The practical discipline is simple: never evaluate a SaaS business on a single number. Decompose MRR into new, expansion, churn, and contraction. Verify every figure against the payment processor. Strip out non-recurring revenue before applying any multiple. Run both ARR and SDE valuations and investigate the gap. If a seller cannot or will not produce component-level MRR data going back 24 months, you have learned something important about either their record-keeping or their honesty, and neither is a reason to proceed.
The good news is that most buyers do not do this work. They see "$480K ARR" and either overpay because the multiple looked reasonable or pass because it looked expensive — without ever knowing which. Doing thirty extra minutes of component analysis on every deal is a genuine, repeatable edge in a market where the majority of participants are reading headlines. If you want that analysis run automatically across new SaaS listings as they hit the market, that is precisely what Deal Alert AI was built to do — surface the deals where the metrics actually support the price, and filter out the ones where they do not. Start with the checklist above on the next three listings you look at, and you will immediately see how many "great SaaS deals" fall apart under thirty minutes of real scrutiny. For more buyer frameworks like this one, browse the rest of the guides at Deal Alert AI.
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.