SaaS Valuation Multiples in 2026

Data current as of August 2026. This page is refreshed quarterly and updated in place — every figure below is dated so you can see how fresh it is.

SaaS valuation multiples in 2026 sit far below where founders remember them. Public SaaS trades around 3.2x trailing revenue. Private M&A clears near 4.0x. Bootstrapped businesses under $1M ARR change hands at 2.5x–4x SDE, which is usually 1x–3x ARR. And AI SaaS multiples split hard by category — from 1x for writing assistants to 10x for healthcare.

Those four numbers describe four different markets, measured on four different bases. Almost all the confusion around SaaS valuation multiples comes from mixing them up — quoting a public revenue multiple at a private business, or an ARR multiple at a company that will be priced on profit. This page keeps every multiple separate, dates each figure, and names its source.

The 2026 numbers at a glance

Everything below is the detail behind those rows: where each multiple comes from, which businesses it actually describes, and what moves it up or down.

Which multiple are you actually looking at?

This is the most expensive misunderstanding in small SaaS, and most published guides to SaaS valuation multiples never address it. A "3x multiple" can mean four completely different prices, because there are four different things it might be multiplying.

Converting between them

The bridge between the SDE world and the ARR world is your profit margin. A business with $120,000 ARR and a 45% SDE margin has $54,000 in owner earnings. At 3.2x SDE that's roughly $173,000 — which is 1.4x ARR, not 3.2x ARR.

So when a founder reads that "SaaS trades at 4x" and applies it to their ARR, they can arrive at a number three times what any buyer will pay. The multiple wasn't wrong. The basis was.

Rule of thumb: below roughly $2M ARR, assume SDE. Between $2M and $5M, buyers run both and take the lower. Above $5M, ARR and EBITDA dominate. For the mechanics of calculating SDE and which costs a buyer will let you add back, see our guide to SaaS business valuation.

Public SaaS multiples in 2026

Public SaaS valuation multiples are the tide, not the benchmark. They set direction for private deals, with a lag of two to four quarters, but they never set your price. A public company has liquidity, analyst coverage, audited financials and a management team. You have none of those, and the discount on your multiple reflects it.

As of Q2 2026, public SaaS trades at a median 3.2x EV/TTM revenue per Software Equity Group — down from 5.7x a year earlier. Aventis Advisors' index of 70 listed SaaS companies put the median at 3.4x in March 2026. PitchBook had enterprise SaaS at 3.3x at the end of Q1 2026.

Public multiples by segment

The single "SaaS" number hides an enormous spread. These are next-twelve-month figures as of 31 August 2026, from Multiples.vc:

DevOps at 8.7x and sales & marketing automation at 1.9x are both "SaaS." That's a 4.5x spread in revenue multiples inside one word, and it is the clearest argument against ever quoting a single industry-average multiple at your own business.

Segment medians: infrastructure SaaS 2.9x, vertical SaaS 2.3x, horizontal SaaS 2.2x. Infrastructure carries the premium because the AI data boom has repriced anything that stores or moves training data, while general-purpose cloud compute drifts toward commodity pricing.

Private SaaS M&A multiples

Private SaaS valuation multiples are what actually matter to a founder, and they behave differently from the public tape — they fall less in a downturn and rise less in a boom, because the buyers setting them are underwriting cash flows rather than trading sentiment.

Aventis Advisors' sample of 543 disclosed private transactions from 2015 to 2026 gives a median of 4.5x EV/revenue, with a first-to-third quartile range of 2.4x to 8.1x. On a profit basis, 232 disclosed deals give a median 23.0x EV/EBITDA, quartiles 12.8x to 47.1x.

Those quartiles matter far more than the median, and almost nobody publishes them. Half of all SaaS deals land outside 2.4x–8.1x. If someone tells you SaaS sells at a 4.5x multiple, they have told you almost nothing about your business — the spread around that multiple is wider than the multiple itself.

Software Equity Group put the Q2 2026 median M&A transaction at 4.0x EV/TTM revenue, down slightly from 4.2x the prior quarter.

The public-to-private discount

Private companies trade at roughly a 20–30% discount to public comparables of similar size and growth. The discount pays for illiquidity, thinner financial reporting, key-person risk and a much smaller pool of buyers. It widens as you get smaller — at the bottom of the market it is far more than 30%, because the buyer pool shrinks to individuals.

Deal volume is rising

Software Equity Group counted 2,784 SaaS transactions in the trailing twelve months to Q2 2026, up 16% year over year. More deals at slightly lower multiples is a specific kind of market: buyers are active and capital is available, but they are disciplined on price. For a seller with clean numbers, that is a better market than low volume at high multiples, because deals actually close.

SaaS valuation multiples by ARR band

Size is the single largest determinant of your multiple, ahead of growth, retention or category. If you only take one table from this page, take this one — SaaS valuation multiples by ARR band is the breakdown most founders are actually searching for.

Sources: the $1M and above bands are Windsor Drake advisory ranges, August 2026. The sub-$1M bands are AIExchange.club's own read from marketplace activity, stated on an SDE basis. We've labelled them as our estimate rather than presenting them as survey data, because no published survey covers this range.

Under $1M ARR — where the published data runs out

This is the gap in every other source on this topic. Windsor Drake's table starts at $1M. Aventis segments by deal size and puts everything under $5M in a single 3.3x bucket. The best coverage anywhere else is a single row: FE International's 2.5x–4x SDE, or Ad Astra's 2x–4x SDE.

Three things are true below $1M ARR that stop being true above it:

• Buyers price profit, not revenue. Almost every deal here is an SDE multiple, because the buyer is purchasing an income stream they will personally operate.
• The buyer pool is individuals. One motivated person with savings, not a firm with a mandate. Demand is thin, seasonal, and easily distracted, and that alone costs half a turn versus the same business at $2M ARR.
• Founder dependency dominates. Above $1M a buyer inherits a team. Below it, they inherit your job. A business needing 30 hours a week of your time is priced as employment, not as an asset.

Why the multiple climbs with size

Aventis's breakdown by deal size shows the effect plainly: deals from $0–5M cleared a median 3.3x, $5–20M 3.2x, $20–50M 3.0x — then $50–100M jumps to 6.1x and $500M+ to 6.2x. The climb is not smooth; it steps up sharply once a business becomes large enough for institutional buyers to compete for it.

Scale buys you competition, and competition is most of your multiple. It also buys durability: a $25M ARR business with a management team survives its founder leaving, while a $250K business often does not.

Want your own number? Get a free SaaS valuation calibrated against real AI SaaS transactions — five questions, about 60 seconds.

AI SaaS valuation multiples by category

Public markets report a single "Artificial Intelligence" line at 4.1x, and no published source we can find breaks AI SaaS valuation multiples down by category. The table below is our own, drawn from the category multiple model behind the AIExchange valuation tool, version-stamped 31 July 2026.

These are ARR multiples for AI SaaS at marketplace scale. Read the scope note under the table before you apply any of them to your business.

How to read this table (and how not to)

These bands describe what a category makes available. They are not a price, and they are emphatically not a promise.

A category band assumes the defensibility that earns it is actually present. Healthcare AI clears 5x–10x because compliance requirements, reimbursement integration and clinical validation are genuinely hard to replicate — when a business really has them. A healthcare-branded AI product at $200K ARR, run by one founder, with 5% monthly churn and no regulatory moat, is not a 7.5x business. It is a 2x business that happens to sit in a 5x–10x category.

The model behind these bands starts at the midpoint and then subtracts, hard, for the things buyers actually price: churn above 3% monthly, more than 25 owner-hours a week, paid acquisition that doesn't pay back, customer concentration, unclear IP ownership. Most businesses land at or below the low end of their category band, and small ones with ordinary metrics almost always do.

So use this table to understand which categories carry structural advantages and which carry structural discounts. Then use the valuation method to work out where your own business actually falls.

Why healthcare and fintech clear the field

Both categories share the same underlying property: switching costs that aren't about software quality. A healthcare AI tool wired into a clinic's billing and records is expensive to remove regardless of how good a competitor is. A fintech product handling money movement carries compliance infrastructure a competitor must rebuild before it can even start.

Buyers pay for that because it converts into retention, and retention is what a multiple is really pricing. A category where customers can leave in an afternoon carries a category-wide discount to its multiple that no amount of current growth will fix.

Why writing and content sit at the floor

Writing assistants are the purest form of the wrapper problem: the product is a prompt and an interface over a model anyone can access. When the model provider ships the same feature natively, the product's reason to exist goes with it.

The retention data bears this out. ChartMogul's analysis of roughly 3,500 software companies found AI-native businesses running 40% gross revenue retention and 48% net revenue retention, against 82% NRR for B2B SaaS overall. Split by price point it gets sharper: AI products under $50/month retained just 23% of revenue, while those above $250/month held 70%. Cheap AI tools churn brutally, and buyers price accordingly.

One honest counterpoint: that same AI-native retention figure improved from 27% to 40% across 2025 as experimental users churned out and committed ones stayed. The category is maturing. If your own retention curve is flattening, that is a documented argument for a higher multiple — but you need the data to show it.

What actually moves a multiple

The Rule of 40 is the largest single premium

Revenue growth rate plus EBITDA margin. Clear 40 and you are, by the market's shorthand, growing efficiently.

PitchBook's Q2 2026 data puts companies clearing the bar at 6.6x trailing revenue against 2.3x for those below it. That is close to a 3x difference in price for a single composite metric. Windsor Drake describes the same effect as one to two extra turns of ARR versus otherwise comparable peers.

The bar is genuinely hard. Aventis found the median Rule of 40 score among public SaaS companies was 28 in Q4 2025, and only about 20% of companies cleared 40. If you clear it, say so early and loudly in any conversation with a buyer. It is the most efficient signal you have.

One caveat for smaller businesses: Rule of 40 is a growth-company metric. A profitable $400K ARR business growing 5% a year with a 50% margin scores 55 and clears the bar comfortably — but no buyer will pay 6.6x revenue for it, because the metric was designed for companies where growth is the investment thesis. Below roughly $1M ARR, treat Rule of 40 as a health check, not a pricing input.

Net revenue retention — and why the benchmarks contradict each other

NRR measures what happens to a cohort's revenue over a year including expansion, contraction and churn. Above 100% means your existing customers grow your revenue without you acquiring anyone.

Buyer thresholds are unusually crisp here. Below 95%, multiples get capped regardless of growth. Between 100% and 110% you are priced at market. Above 115%, expansion revenue compounds at almost no acquisition cost, and that opens a two-to-three turn spread against sub-95% businesses.

Now the part nobody explains. Depending on which report you open, the median NRR for B2B SaaS is 82% or 103%. Both figures are correct, and they measure different populations:

If you run a self-serve product at $29/month, 82% is your benchmark and 103% is a fantasy imported from companies with account managers and annual contracts. If you sell $2,000/month contracts to mid-market buyers, the reverse is true. Quoting the wrong one at yourself — or at a buyer — is how founders end up defending a number that was never theirs.

SaaS Capital's same 2026 survey found bootstrapped growth had slowed to a 15% median, down from 20% the prior year, with gross revenue retention holding at 91%. Growth is compressing across the board; retention is not.

Gross margin and the 70% line

Pure software runs 75–85% gross margin, and buyers need to see that to justify pricing on a revenue multiple at all. Below 70%, the conversation shifts to EBITDA, because the business is starting to look like a services company with software attached — and services companies trade at half the multiple.

This is now the central AI SaaS problem. Inference is a variable cost that scales with usage, so it belongs in cost of goods sold, and a16z's analysis puts AI-heavy businesses at 50–60% gross margins against the traditional 60–80%+ benchmark. If your margin is thin because of model spend, expect detailed questions about routing, caching and smaller models. Inference pricing has been falling roughly 10x a year, so the problem may solve itself — but you'll need to show the trend, not assert it.

How buyer type changes the number

The same business gets materially different offers depending on who is looking at it.

Strategic buyers pay the most because they're buying something other than cash flow — a customer list they can cross-sell, a capability that would take two years to build. Serial consolidators like Constellation Software and Banyan Software pay disciplined prices but close reliably. Individual buyers pay least and are hardest to close.

Bootstrapped and funded businesses trade differently

A bootstrapped business is usually profitable, slower-growing, and priced on earnings. A venture-backed business is usually unprofitable, faster-growing, and priced on revenue — but carries liquidation preferences that can mean the founders see very little of a headline number. Two companies at identical ARR can produce completely different outcomes for the person who built them.

How multiples got here: 2021 to 2026

Figures from Aventis Advisors, PitchBook and Software Equity Group. The two end-2025 readings differ because the samples differ — Aventis tracks 70 companies with $1B+ market caps, PitchBook a broader enterprise set.

The 2021 peak was an artefact of zero interest rates, and it is not coming back on any timeline worth waiting for. SaaS valuation multiples have fallen roughly 84% from that high. The more useful comparison is the pre-2020 baseline of about 5x–7x for public SaaS, which is where multiples were heading before the distortion — and today's 3.2x sits below even that, reflecting genuine uncertainty about what AI does to incumbent software.

The practical read for a founder: waiting for a better market is not a plan. Deal volume is up 16% year over year, which means buyers are active now. Improving your own retention, concentration and founder-dependency will move your multiple far more, and far faster, than any plausible market recovery.

Look up your own multiple

Pick your ARR band, category and growth profile to see the applicable multiple range and where you'd sit within it. The lookup reads the same category data as the tables above, so the multiples here can never drift apart from the ones we publish.

Get a full free valuation → Five questions, about 60 seconds, no credit card, no email required to see your range.

Methodology and sources

Every multiple on this page is attributed and dated. Where we publish our own estimate rather than someone else's data, we say so plainly.

• Public multiples — Software Equity Group (Q2 2026), Aventis Advisors SaaS Index (70 listed companies, March 2026), PitchBook (Q1 2026), Multiples.vc segment data (31 August 2026, NTM basis, underlying data from FactSet and Morningstar).
• Private M&A multiples — Aventis Advisors, 543 disclosed transactions with revenue multiples and 232 with EBITDA multiples, 2015–2026. Disclosed deals over-represent larger, cleaner transactions; the unreported long tail of small deals prices lower.
• ARR bands $1M and above — Windsor Drake advisory ranges, August 2026.
• Sub-$1M ARR bands — AIExchange.club's own read from marketplace activity, cross-checked against FE International (2.5x–4x SDE) and Ad Astra Equity (2x–4x SDE). Directional, not a statistical sample.
• AI SaaS category bands — AIExchange.club's category multiple model, version 2026-07-31, built from marketplace-scale AI SaaS comparables. Bands are pre-adjustment; realised multiples are typically lower.
• Retention data — ChartMogul (~3,500 companies, data through September 2025) and SaaS Capital (1,000+ private B2B SaaS, 2026).
• AI gross margins — a16z analysis via Avante Ventures.

Updated quarterly. If you spot a figure that has moved, tell us and we'll correct it.

Frequently asked questions

What is the average SaaS valuation multiple in 2026?

Public SaaS trades at a median 3.2x trailing revenue and private M&A clears around 4.0x as of Q2 2026. But average SaaS valuation multiples are close to useless as a guide — the first-to-third quartile range on private deals is 2.4x to 8.1x, so half of all transactions fall outside it. Your size, retention and growth move your multiple far more than the market average does.

What is a good ARR multiple for a SaaS business?

It depends almost entirely on size. Under $1M ARR, most businesses sell on SDE at 2.5x–4x, which usually works out to 1x–3x ARR. At $1M–$3M ARR, 2.5x–4.0x is typical. At $10M–$25M, 4.0x–6.5x. Anything above 5x ARR below $3M requires exceptional retention, growth or a strategic buyer.

Why are SaaS multiples lower than they were in 2021?

Public SaaS peaked near 20x revenue in late 2021 under zero interest rates and a growth-at-any-cost consensus. Rates rose, capital got expensive, and buyers started pricing profitability instead. Multiples fell to 6.7x by early 2023 and to 3.2x by 2026 as AI uncertainty added a further discount to incumbent software. The 2021 level was the anomaly, not the baseline.

What EBITDA multiple do SaaS companies sell for?

Aventis Advisors' sample of 232 disclosed private transactions gives a median 23.0x EV/EBITDA, with quartiles from 12.8x to 47.1x. Public listed software with positive EBITDA traded around 24.5x in early 2026. EBITDA multiples only apply to businesses with real management in place — owner-operated companies are valued on SDE instead.

How does the Rule of 40 affect valuation multiples?

It is the largest single premium available. PitchBook's Q2 2026 data shows companies clearing 40 trading at 6.6x trailing revenue versus 2.3x for those below — nearly a 3x difference. Only about 20% of SaaS companies clear the bar. Below roughly $1M ARR it matters much less, because buyers there are pricing owner earnings rather than growth efficiency.

Do AI SaaS businesses get higher multiples than traditional SaaS?

Some do, most don't. AI businesses with proprietary data, workflow lock-in or regulatory moats — healthcare and fintech especially — clear 5x ARR and up. Thin wrappers over a foundation model trade at 1x–3x, and the retention data explains why: AI-native companies average 48% net revenue retention against 82% for B2B SaaS generally, dropping to 23% gross retention for products under $50 a month.

What multiple do micro-SaaS businesses under $1M ARR sell for?

2.5x–4x SDE is the working range, which typically lands between 1x and 3x ARR. Under $250K ARR expect 2.0x–3.2x SDE. Buyers at this size are individuals rather than firms, so the pool is thin and founder dependency is the biggest single discount. Almost no published survey covers this range, so treat any precise-sounding figure — including ours — as directional.

Why do public and private SaaS multiples differ?

Private companies trade at roughly a 20–30% discount to public comparables, widening as they get smaller. The discount pays for illiquidity, less rigorous financial reporting, key-person risk and a much smaller buyer pool. Private multiples also lag public ones by two to four quarters, so a public re-rating shows up in private deals months later.

What net revenue retention do buyers expect in 2026?

Below 95% caps your multiple regardless of growth; 100–110% is market rate; above 115% opens a two-to-three turn premium. But the benchmark depends on your model: median NRR is 82% across ~3,500 mostly self-serve companies (ChartMogul) and 103% among bootstrapped B2B firms at $3M–$20M ARR (SaaS Capital). Use the one that matches how you actually sell.

Are SaaS multiples expected to recover?

Nobody credibly knows, and the 2021 peak specifically is not returning — it depended on zero interest rates. The pre-2020 baseline of 5x–7x for public SaaS is a more reasonable ceiling to imagine for a recovery in multiples. Meanwhile deal volume rose 16% year over year to 2,784 transactions, so buyers are active at today's multiples. Improving retention, concentration and founder dependency will move your own multiple faster than waiting for the market to move it for you.

Working out what your own business is worth? Start with how SaaS business valuation works, or browse live AI SaaS listings to see what comparable businesses are actually priced at.