B2B SaaS valuation multiples: 2021 to 2026
What software companies are actually selling for right now, from the public markets down to sub-$10M founder exits: sourced, dated, and refreshed quarterly. Every chart on this page is also published as a plain data table, and every figure carries its publisher and as-of date. Read it before a buyer quotes you a number.
Multiples reset hard, and the public premium is nearly gone
Public SaaS fell from an 18x to 19x revenue peak to 3.4x. Private deals never flew that high, so they fell far less. The two now trade within 0.3x of each other.
View the data as a table
| Period | Median EV/Revenue |
|---|---|
| 2021 | 18-19x (peak) |
| 2023 | 6.7x |
| early 2025 | 7.3x |
| Jul 2025 | 6.0x |
| Mar 2026 | 3.4x (latest) |
View the data as a table
| Period | Median EV/Revenue |
|---|---|
| 2021 | 6.4x (peak) |
| 2023 | 3.3x |
| 2024 | 2.9x (low) |
| 2025 | 3.8x |
| Q1 2026 | 3.1x (latest) |
| Long-run median, 2015-2026 | 4.5x (middle half 2.4x to 8.1x) |
Source: Aventis Advisors, median EV/Revenue of public SaaS companies (left); Aventis Advisors, median EV/Revenue of private SaaS M&A transactions (right). Both panels share one scale. 2022 is not published as a standalone figure in the verified vintage. The sharp public-market step from mid-2025 to March 2026 is as published by Aventis and is corroborated by independent series: SEG’s index median printed 3.2x in Q2 2026 and Clouded Judgement’s 3.6x in February 2026 was a ten-year low.
other publishers 3.2xSEG public SaaS index, median EV/TTM revenue (Q2 2026)3.6xClouded Judgement, median EV/NTM revenue, lowest in 10+ years (Feb 2026)4.0xSEG, median SaaS M&A transaction EV/TTM revenue (Q2 2026)
The arc from 2021 to 2026, in four moves
2021: the peak that never reached private deals. Listed SaaS companies traded at a median of 18x to 19x revenue for most of 2021 (Aventis Advisors). Private SaaS transactions peaked at a 6.4x median the same year. That gap is the first honest lesson of this report: even at the top of the cycle, the median private seller collected roughly a third of the public print. The public number is the one that gets quoted in pitch decks; the private one is the one that gets paid.
2022 to 2023: the rate reset. As the cost of capital repriced, the public median fell to 6.7x by 2023 and the private median to 3.3x. Both markets roughly halved and then halved again from their peaks. Sellers who anchored to 2021 comps spent these two years declining offers that, in hindsight, were fair prints for the new market.
2024 to 2025: volume returned before price did. The private median bottomed at 2.9x in 2024, then recovered to 3.8x across 2025. Meanwhile deal count set an all-time record: 2,698 closed SaaS transactions in 2025, +28% on the 2,107 closed in 2024 (SEG), and $362bn of global software M&A value (Silverpeak). Strategic and private equity buyers came back in force; they simply came back disciplined.
2025 to 2026: the public premium evaporated. The strangest stretch in the dataset is the most recent one. While private medians held near 3.1x into Q1 2026, the public median stepped down from 6.0x in July 2025 to 3.4x by March 2026 (Aventis). That step is not a quirk of one index: SEG’s public SaaS median printed 3.2x in Q2 2026, and Clouded Judgement’s 3.6x EV/NTM reading in February 2026 was the lowest in more than ten years. Public and private software now price within 0.3x of each other.
Our reading, clearly labeled as interpretation: the market has stopped paying for the category and gone back to paying for the company. When an index median compresses this far while deal volume sets records, the money has not left; it has become selective. That is a hostile market for an unprepared seller and a workable one for a positioned seller, which is the recurring theme of the sections below.
What is the average SaaS valuation multiple in 2026?
As of the latest verified data, the median public SaaS company trades at 3.4x EV/Revenue (Aventis Advisors, March 2026) and the median private SaaS acquisition prices at 3.1x EV/Revenue (Aventis, Q1 2026). If you hold one number in mind for 2026, hold roughly 3x revenue for a median B2B SaaS business.
Two cautions before you use that number. First, medians, not averages, describe the market you will actually meet: averages are dragged upward by a small set of premium deals. Second, dispersion is enormous. Across Aventis’s 2015 to 2026 private dataset the long-run median is 4.5x, but the middle half of deals ranges from 2.4x to 8.1x, and disclosed outliers in section 06 cleared far above that. “The market multiple” is a starting point for a negotiation, not a valuation of your company.
ARR, EBITDA or SDE: the basis matters more than the headline
A “5x” deal and a “5x” deal can be wildly different prices. Before comparing any two multiples, establish what they multiply, over which period, and from which population of companies.
Revenue multiples (EV/Revenue, often quoted on ARR). The standard basis for growing software companies, and the basis of every trend chart in section 01. Public series are typically enterprise value over trailing-twelve-month (TTM) or next-twelve-month (NTM) revenue; private deals are more often quoted on annual recurring revenue (ARR). TTM, NTM and ARR can differ by 20% or more for a fast-growing company, which is why every figure on this page states its basis. Revenue multiples dominate when the buyer is underwriting growth rather than current profit.
EBITDA multiples. The profit basis for established companies. Per FE International’s published guidance, once a business is worth more than roughly $5M, EBITDA multiples are used almost exclusively. A buyer paying on EBITDA is underwriting cash generation, so quality of earnings, margin durability and owner-independence drive the number.
SDE multiples. Seller discretionary earnings is EBITDA with the owner’s compensation and discretionary costs added back, and it is the standard basis below roughly $5M of value (FE International, January 2026). For an owner-operated business, SDE is usually meaningfully larger than EBITDA, so an SDE multiple will look smaller than the equivalent EBITDA multiple for the same cheque.
The same cheque, three headlines. A purely arithmetic illustration, not market data: a company with $3M of ARR and $900K of EBITDA that sells for $9M has sold at 3.0x revenue and 10x EBITDA simultaneously. Neither number is more true; they describe the same price. This is how a founder reading marketplace SDE data and a founder reading public EV/Revenue data can both feel the other’s numbers are wrong.
Enterprise value is not your wire transfer. Multiples in this report are enterprise value based: cash-free, debt-free, and before deal structure. Earnouts, escrows, rolled equity and working-capital adjustments all sit between the headline multiple and the money that lands. A structured 5x can be worth less than a clean 4x.
When anyone quotes you a multiple, ask three questions: a multiple of what (revenue, EBITDA or SDE), over which period (TTM, NTM or ARR), and measured on which population (a public index, closed private deals, or asking prices). Every figure on this page answers all three; most numbers quoted in inboxes answer none.
Below institutional scale, profit multiples set the price
Institutional revenue-multiple logic fades below a few million dollars of value. FE International puts the switch at roughly $5M: below it, deals price on SDE; above it, on EBITDA. Preparation moves the number more than the market does.
What multiple can a company under $10M expect?
Survey medians and published guidance, clearly separated: private bootstrapped B2B SaaS companies reported a median of 4.8x ARR at the start of 2025, with equity-backed peers at 5.3x (SaaS Capital). Brokered sales guidance runs wider, at 4x to 10x SDE (FE International).
The table below separates the two kinds of evidence. Survey medians (SaaS Capital) describe what companies actually reported. FE International’s ranges are vendor-published brokerage guidance: honest about what a well-run process can achieve, but guidance nonetheless, and we label them as such.
| Segment | Basis | Multiple | Source | As of |
|---|---|---|---|---|
| Businesses valued under $2M | SDE | 5.0x to 7.0x | FE International, vendor-published guidance | January 2026 |
| Businesses valued over $2M | SDE | 7.0x to 10.0x | FE International, vendor-published guidance | January 2026 |
| Bootstrapped private B2B SaaS | ARR | 4.8x median | SaaS Capital survey | start of 2025 |
| Equity-backed private B2B SaaS | ARR | 5.3x median | SaaS Capital survey | start of 2025 |
FE International figures are published brokerage guidance (vendor-published ranges), not closed-deal medians. SaaS Capital figures are survey-observed medians.
One operating truth runs through the whole segment. Buyers at this end of the market are buying current cash flow, and the sellers who win are the ones who arrive with clean books, documented margins, and revenue that does not depend on the founder answering support tickets. Prepared processes clear faster than unprepared ones, and the preparation is entirely on the seller’s clock.
Geography moves the multiple: 5.0x in the US, 2.6x in Germany
Median EV/Revenue of private software M&A by target country. The demand around your deal prices it; the local market is rarely where that demand peaks.
View the data as a table
| Target country | Median EV/Revenue | Deals in sample |
|---|---|---|
| United States | 5.0x | 467 |
| United Kingdom | 3.3x | 179 |
| Sweden | 3.0x | 51 |
| France | 2.8x | 47 |
| Germany | 2.6x | 38 |
US public SaaS traded 20% to 40% above European peers through 2024 (GP Bullhound). By early 2026, for the first time since 2015, US SaaS no longer trades at a premium to global SaaS in public markets (Aventis).
European buyers took 58% of European-target software M&A by value in late 2025, up from 42% over the prior two years; North American buyers still account for 67% of European-target deal count (Silverpeak).
This is why our processes are global by default. The best buyer for a European software company is usually not a domestic one; 80% of Array Capital transactions close cross-border.
See where our deals closed →Source: Aventis Advisors, median EV/Revenue of private software M&A by target country, 2015-2025 dataset.
Why are European SaaS multiples lower?
Because the deepest pool of software buyers is American, and historically it paid a premium at home: the median US software target cleared 5.0x revenue across 467 deals in the 2015 to 2025 dataset, against 3.3x in the UK and 2.6x in Germany (Aventis). Thinner domestic strategic demand and historically discounted European public comps compound into lower private medians.
Two things complicate the simple story, and both are in the data. First, the public-market side of the discount has closed: as of early 2026, for the first time since 2015, US SaaS no longer trades at a premium to global SaaS in public markets (Aventis), after running 20% to 40% above European peers through 2024 (GP Bullhound). Second, the buyer mix in Europe is shifting: European buyers took 58% of European-target software M&A by value in late 2025, up from 42% over the prior two years, while North American buyers still account for 67% of European-target deal count (Silverpeak).
The practical conclusion for a European founder is not that a discount is destiny. Country medians measure where deals happened, not where they had to happen. A process that puts US strategics, pan-European consolidators and global private equity in the same competitive tension prices the company off the deepest demand it can reach, not off its postcode. That is the entire argument for running a structured, cross-border process rather than responding to the one inbound that found you.
Retention and growth, not size, decide your premium
The spread between a discounted exit and a premium one is measured and public. Two datasets, two lenses, one conclusion: buyers pay for durable revenue.
View the data as a table
| Cohort | EV/TTM revenue |
|---|---|
| NRR above 120% | 11.7x |
| Index median | 5.6x |
| NRR below 100% | 4.1x |
Same dataset: Rule of 40 above 40% commands 12.4x; gross margin above 80% trades at 7.6x versus 5.5x below it. Note: these are Q4 2024 public-market levels; absolute multiples have compressed since (section 01). The premium ratios, not the levels, are the point.
View the data as a table
| Cohort | EV/NTM revenue |
|---|---|
| Growth above 22% | 9.7x |
| Growth 15% to 22% | 6.4x |
| Growth below 15% | 2.6x |
These are public-market lenses; private buyers apply the same logic with a size discount. We position every mandate on these drivers before it goes to market.
What moves a company from a median multiple to a top-quartile one?
Retention and growth, in that order of controllability. Companies with net revenue retention above 120% carried an 11.7x EV/TTM revenue multiple against an index median of 5.6x (SEG, Q4 2024), and software growing above 22% traded at 9.7x EV/NTM revenue against 2.6x for sub-15% growers (Clouded Judgement, February 2026). Between the bottom and top cohorts, that is roughly a 3x to 4x difference in price for the same dollar of revenue.
The same SEG dataset shows the supporting cast: a Rule of 40 score above 40% commanded 12.4x, and gross margin above 80% traded at 7.6x versus 5.5x below it. None of these are size metrics. In every panel we can verify, a smaller company with durable revenue outprices a larger one with leaky revenue.
For a founder planning an exit, the sequencing implication is concrete: the multiple is mostly set before the process starts. Churn fixed two quarters before going to market shows up in the numbers a buyer diligences; churn fixed during diligence reads as a story. This is why we start mandate preparation on these drivers, not on the deck.
What this means if you are selling in the next 12 months
The data above is only useful if it changes what you do. Five conclusions we would act on, in order.
1. Stop waiting for the old market. Nothing in the 2021 to 2026 arc suggests a return to 18x public medians, and the private market never paid them anyway. What 2025 and 2026 do show is record transaction volume: 2,698 closed SaaS deals in 2025 (SEG). Buyers are active and funded; they are simply pricing at roughly 3x median revenue with a wide, earnable spread above it. A good company with a good process sells well in this market today.
2. Know your number on all three bases. Before the first buyer call, know your TTM revenue, ARR, EBITDA and SDE, and what the relevant multiple on each basis implies for your range. Sub-$5M sellers should expect SDE conversations (FE International); larger sellers should expect EBITDA and ARR framing. Sellers who cannot bridge between bases get anchored by whoever quotes first.
3. Spend the next two quarters on retention, not on the deck. The measured premium for NRR above 120% versus below 100% is 11.7x against 4.1x (SEG). No positioning document moves price like a retention fix that has had time to show up in the cohort data a buyer will diligence.
4. Run the process wide and cross-border. Country medians in section 04 differ by nearly 2x. The single highest-leverage process decision, especially for European sellers, is whether the buyer list is global. One inbound acquirer is a price-taker's position; five at the table is a price-setter's.
5. Prepare like the timeline is a variable, because it is. Preparation is most of the variance we see in practice: clean financials, documented margins, transferable operations, and answers ready before the questions arrive. The route-by-route timeline data is in our report on how long it takes to sell a software company.
If you want the operational version of this list, we keep a full playbook at how to sell a SaaS company, and a dedicated guide for selling a Shopify app. If you would rather talk through where your company sits against these benchmarks, a discovery call is free and stays confidential.
How this report is built
Most multiple roundups average other people's averages. This one does not, and the restraint is deliberate: the honesty of the labels is what makes the numbers usable in a negotiation.
- One publisher series per chart. Index constructions differ (the 2021 public peak ranges from about 17x to 28x across publishers depending on weighting and sample), so we never mix methodologies inside a figure. A blended line would look more authoritative and mean less.
- Where publishers conflict, we cite the series we verified directly at the primary source and note the disagreement rather than averaging it away. An average of two incompatible methodologies describes neither market.
- Every number carries its as-of date and appears exactly as published. Figures that are broker guidance rather than closed-deal data are labeled vendor-published.
- Every chart is also published as a plain HTML table with the same figures, so the data is readable without rendering a single pixel.
- Updated quarterly in place at this URL. First edition: July 2026; each refresh is logged here. Figures reflect data available as of July 2026.
- Nothing on this page is investment advice, and no multiple prices a specific company. Your company trades on its own facts; that is the whole point of positioning.
Considering an exit in the next 12 to 24 months?
A 30-minute discovery call: our read on your valuation, buyer universe, and timing. No commitment, full confidentiality.