SaaS Metrics Benchmarks 2026: CAC LTV NRR by Stage
Comprehensive 2026 SaaS benchmark reference: CAC, LTV, NRR, churn, Rule of 40 benchmarks by funding stage and ACV segment, with AI-native SaaS context.
Whether you're looking for an angel investor, a growth advisor, or just want to connect — I'm always open to great ideas.
Get in TouchAI, startups & growth insights. No spam.
TL;DR: SaaS benchmark expectations have permanently shifted since the 2022 correction. This is the complete 2026 reference for CAC, LTV, NRR, churn, burn multiple, and Rule of 40 organized by funding stage and ACV segment — including how AI-native SaaS is bending the traditional metrics framework.**
I review dozens of pitch decks and investor update emails every month. The single most common mistake I see is founders citing benchmarks incorrectly — either to make their metrics look better than they are, or to dismiss underperformance by claiming the benchmark does not apply to them.
Both errors are costly. The first misleads investors into a deal that will disappoint. The second prevents a founder from identifying a real problem early enough to fix it.
Benchmarks serve three legitimate purposes, and only three:
1. Diagnostic signals: When your metric is significantly below benchmark for your stage, it is a signal to investigate, not a verdict of failure. A CAC payback period of 36 months is a warning light, not a death sentence. Maybe your customer lifetime is exceptionally long and the LTV math still works. Maybe you are in an enterprise segment where 36 months is actually competitive. The benchmark tells you to look harder, not to panic.
2. Fundraising context: Investors use benchmarks to calibrate expectations. A Series A company with NRR below 90% faces a different conversation than one with NRR of 115%. Understanding the benchmark helps you either defend why your deviation is intentional and understood, or acknowledge the gap and explain the improvement plan.
3. Hiring and headcount decisions: Sales efficiency benchmarks help you determine whether hiring another sales rep is likely to be accretive or dilutive. If your Magic Number is 0.4, adding sales headcount is probably destroying value until you fix the underlying efficiency problem.
What benchmarks do not do: they do not tell you what to optimize for. Optimizing NRR at the expense of new logo growth may produce a beautiful NRR number and a dying company. Every metric exists in context. A high-NRR, low-growth company in a winner-take-all market is in trouble. A low-NRR, high-growth company with improving cohort retention is probably fine.
The correct use of a benchmark is: "This metric is below benchmark. Is that because of a problem or because of a deliberate strategic choice? If it is a problem, what is causing it? If it is a choice, what are we getting in exchange for the underperformance on this metric?"
The benchmarks in this post are compiled from the following primary sources, all published between Q3 2024 and Q1 2026:
| Source | Coverage | Methodology |
|---|---|---|
| OpenView Partners SaaS Benchmarks 2025 | 600+ SaaS companies, primarily Series A-C | Annual survey + proprietary portfolio data |
| SaaStr Annual 2025 Benchmarking Data | 1,500+ SaaS leaders | Conference survey, self-reported |
| a16z SaaS Metrics Guide 2025 | Proprietary portfolio benchmarks | Portfolio company analysis |
| Bessemer Venture Partners State of the Cloud 2025 | Public + private SaaS benchmarks | Public market analysis + BVP portfolio |
| CB Insights State of Venture Q4 2025 | Market-level data | Deal flow and exit analysis |
| KeyBanc Capital Markets SaaS Survey 2025 | 350+ private SaaS companies | Annual survey |
| Sapphire Ventures SaaS Efficiency Report 2025 | Series A-D benchmarks | Proprietary portfolio + survey |
A note on self-reported data: survey benchmarks are subject to selection bias. Companies performing well are more likely to participate. Treat the benchmarks in this post as directional guidance, particularly for the early-stage figures where sample sizes are smaller and variance is high.
I have cross-referenced figures across multiple sources where possible. Where sources conflict significantly, I have noted the range rather than picking a single number.
All figures in this post reflect the post-2022 efficiency era. Benchmarks from 2020-2021 are not applicable to the current fundraising environment and I have excluded them deliberately.
Customer Acquisition Cost (CAC) is the fully loaded cost to acquire one new customer, including sales salaries, marketing spend, and overhead allocated to new customer acquisition. The most common error in CAC calculation is excluding founder time, SDR compensation, or marketing tool costs. If your CAC looks suspiciously low, check your inputs.
| Stage | Median CAC | Top Quartile CAC | Bottom Quartile CAC | Notes |
|---|---|---|---|---|
| Pre-seed / Seed | $1,500 - $8,000 | < $2,000 | > $15,000 | Often founder-led; low marketing spend |
| Series A | $5,000 - $25,000 | < $8,000 | > $40,000 | First real sales motion; CAC rising |
| Series B | $15,000 - $60,000 | < $20,000 | > $80,000 | Scaling sales org; efficiency test |
| Series C+ | $25,000 - $100,000+ | < $30,000 | > $150,000 | Depends heavily on ACV segment |
ACV segment is a stronger predictor of CAC range than funding stage, because it reflects the complexity and length of the sales process.
| ACV Segment | Typical CAC Range | Median CAC | Sales Motion |
|---|---|---|---|
| SMB (ACV < $5k) | $300 - $3,000 | $800 | Product-led; low-touch; high volume |
| Mid-market (ACV $5k-$50k) | $5,000 - $30,000 | $12,000 | Inside sales; moderate cycle |
| Enterprise (ACV $50k-$200k) | $30,000 - $100,000 | $55,000 | Field sales; long cycle; procurement |
| Strategic Enterprise (ACV $200k+) | $80,000 - $300,000+ | $150,000 | Multi-stakeholder; 9-18 month cycle |
At seed, CAC is often artificially low because the founders are closing deals personally. Founder-led sales does not scale, but it does compress CAC. As companies build sales teams, CAC rises. The inflection point where CAC stabilizes — or ideally begins to decline through brand, word of mouth, and improved sales process — is one of the clearest signals of whether a go-to-market is working.
Blended CAC also masks important segment-level differences. A company with $15,000 median CAC may have $2,000 CAC from product-led channels and $40,000 CAC from enterprise field sales. These represent fundamentally different business models operating inside the same company and should be benchmarked separately.
Lifetime Value (LTV) is the total net revenue expected from a customer over their relationship with the company. The most common calculation is: LTV = (ARPU × Gross Margin) / Customer Churn Rate.
The LTV:CAC ratio is the primary efficiency metric for SaaS go-to-market. It tells you how much lifetime value you generate for each dollar spent acquiring a customer.
| Stage | Healthy LTV:CAC | Excellent LTV:CAC | Warning Zone | Notes |
|---|---|---|---|---|
| Pre-seed / Seed | 2x - 4x | > 4x | < 1.5x | Low sample size; projections dominate |
| Series A | 3x - 5x | > 5x | < 2.5x | Investors expect improving trajectory |
| Series B | 4x - 7x | > 7x | < 3x | Scale should improve ratio |
| Series C+ | 5x - 10x | > 10x | < 4x | Mature GTM; high bar for efficiency |
| ACV Segment | Typical LTV:CAC | Notes |
|---|---|---|
| SMB | 2x - 4x | High churn compresses LTV |
| Mid-market | 3x - 6x | Better retention, moderate CAC |
| Enterprise | 5x - 12x | Long lifetime, high CAC offset by low churn |
A healthy LTV:CAC does not mean you should invest more in customer acquisition immediately. It means the unit economics of your acquisition engine are working. Whether to increase investment depends on whether you can deploy capital efficiently at higher volume — which is a question of go-to-market capacity, not LTV:CAC alone.
Using projected LTV instead of cohort-derived LTV. Projected LTV is a guess. Cohort-derived LTV is a measurement. Until you have 24+ months of customer data, your LTV is an estimate, and you should be explicit about that in investor conversations.
Using nominal revenue instead of gross-margin-adjusted LTV. A company with 50% gross margins has half the LTV of a company with 80% gross margins at the same ARPU and churn rate. Gross margin must be in the calculation.
Using average churn instead of cohort churn. If your recent cohorts have better retention than older cohorts, using blended average churn underestimates LTV. If recent cohorts have worse retention, it overestimates LTV. Always segment by cohort.
CAC payback period measures how long it takes to recover the cost of acquiring a customer through gross profit from that customer. It is calculated as: CAC Payback = CAC / (MRR × Gross Margin).
Shorter payback periods mean faster capital recycling, which is particularly important for capital-efficient SaaS businesses that want to grow without continuously raising equity.
| Stage | Best-in-Class | Good | Acceptable | Concerning |
|---|---|---|---|---|
| Seed | < 12 months | 12-18 months | 18-24 months | > 24 months |
| Series A | < 12 months | 12-18 months | 18-24 months | > 24 months |
| Series B | < 12 months | 12-18 months | 18-24 months | > 30 months |
| Series C+ | < 12 months | 12-18 months | 18-24 months | > 24 months |
| ACV Segment | Median Payback | Best-in-Class | Why It Varies |
|---|---|---|---|
| SMB | 6 - 14 months | < 8 months | Low CAC; lower margins; higher churn |
| Mid-market | 12 - 22 months | < 15 months | Moderate CAC; good margins |
| Enterprise | 18 - 36 months | < 24 months | High CAC; high margins; very low churn |
Before the 2022 market correction, investors routinely funded companies with 36-48 month payback periods on the assumption that fast growth would eventually produce efficiency. That assumption is no longer operative at most growth stages.
In the current environment, Series A and Series B investors expect CAC payback periods below 24 months as a baseline. Companies with payback periods above 24 months face significant pressure to either reduce CAC or increase ACV before they can raise their next round at a reasonable valuation.
The exception is pure enterprise businesses with ACV above $200,000 and very low churn. In that segment, 24-36 month payback is accepted because the long customer lifetime and expansion revenue make the lifetime economics compelling even with extended payback periods.
Net Revenue Retention (NRR) measures whether your existing customer base is growing or shrinking in revenue terms, independent of new customer acquisition. It includes expansion revenue (upsells, seat additions, tier upgrades), contraction (downgrades), and churn.
Gross Revenue Retention (GRR) is NRR without expansion revenue — it measures pure retention before accounting for upsell. GRR cannot exceed 100%.
NRR above 100% means your existing customer base grows revenue over time even if you acquire zero new customers. For a capital-constrained SaaS business, NRR above 110% is an extraordinary advantage — it means growth is partially self-funded through existing customer expansion.
The formula: NRR = (Beginning MRR + Expansion MRR - Churned MRR - Contracted MRR) / Beginning MRR
| Stage | Best-in-Class | Strong | Good | Below Average | Concerning |
|---|---|---|---|---|---|
| Seed | > 110% | 105-110% | 100-105% | 90-100% | < 90% |
| Series A | > 115% | 110-115% | 105-110% | 95-105% | < 95% |
| Series B | > 120% | 115-120% | 110-115% | 100-110% | < 100% |
| Series C+ | > 125% | 120-125% | 115-120% | 105-115% | < 105% |
| ACV Segment | Median NRR | Top Quartile NRR | Bottom Quartile NRR |
|---|---|---|---|
| SMB (< $5k ACV) | 92% - 98% | > 105% | < 85% |
| Mid-market ($5k-$50k ACV) | 105% - 115% | > 120% | < 95% |
| Enterprise ($50k-$200k ACV) | 115% - 125% | > 130% | < 105% |
| Strategic Enterprise ($200k+ ACV) | 120% - 135% | > 140% | < 110% |
The SMB segment structurally produces lower NRR because smaller companies have higher closure rates (customers go out of business), lower average contract values limit upsell room, and the customer base turns over faster. An SMB-focused SaaS with NRR of 100% may be performing comparably to an enterprise SaaS with NRR of 120% once the structural differences are accounted for.
| ACV Segment | Best-in-Class GRR | Good GRR | Acceptable GRR |
|---|---|---|---|
| SMB | > 85% | 80-85% | 75-80% |
| Mid-market | > 90% | 87-90% | 83-87% |
| Enterprise | > 95% | 92-95% | 88-92% |
GRR is particularly important for understanding the health of the core business separate from upsell motion. A company with 90% NRR but 75% GRR is masking significant churn with heavy upsell. That pattern is unsustainable if the churn problem is not addressed — eventually you run out of customers to upsell.
ARR growth rate and the Rule of 40 are the two most commonly referenced performance metrics in fundraising conversations at Series A and beyond.
| Stage | Top Quartile Growth | Median Growth | Bottom Quartile Growth | Notes |
|---|---|---|---|---|
| Seed to Series A | > 300% YoY | 150-300% | < 100% | T2D3 target: 3x |
| Series A | > 200% YoY | 100-200% | < 80% | Investor bar has risen since 2022 |
| Series B | > 100% YoY | 60-100% | < 50% | Efficiency increasingly weighted |
| Series C | > 60% YoY | 40-60% | < 30% | Rule of 40 starts to matter heavily |
| Growth / Pre-IPO | > 40% YoY | 25-40% | < 20% | Rule of 40 / 60 dominant lens |
The T2D3 framework — triple, triple, double, double, double — describes the growth trajectory that historically produced venture-scale outcomes in SaaS:
A company that executes T2D3 starting from $2M ARR reaches approximately $150M ARR by year six. At a 10x revenue multiple, that produces a $1.5B valuation. This is the model that shaped a decade of SaaS venture investment.
T2D3 is harder to achieve in 2026 than it was in 2019. The SaaS market is more competitive, customer acquisition costs are higher, and growth-at-all-costs capital is not available. Fewer than 15% of Series A SaaS companies hit T2D3 through their first two growth years in 2024-2025. The benchmark is still relevant as an ambition but should not be used as a baseline expectation.
The Rule of 40 states that growth rate plus profit margin should equal or exceed 40%. It is the primary metric for evaluating the efficiency of a scaled SaaS business.
| Stage | Rule of 40 Score | Interpretation |
|---|---|---|
| Series B | > 40 | Excellent; likely fundraising from a position of strength |
| Series B | 30-40 | Good; improvable; investors will want trajectory |
| Series B | 20-30 | Acceptable; pressure to improve |
| Series B | < 20 | Concerning; requires explanation |
| Series C+ | > 50 | Exceptional |
| Series C+ | 40-50 | Strong |
| Series C+ | 30-40 | Acceptable |
| Series C+ | < 30 | Below bar; IPO prospects complicated |
A new benchmark is emerging for high-gross-margin AI-native SaaS companies: the Rule of 60. AI products that have gross margins above 80% and strong NRR can sustain profitability at lower growth rates, but they can also demonstrate exceptional efficiency ratios that make a Rule of 60 target reasonable.
Several AI infrastructure and AI-native SaaS companies that went public or received late-stage valuation marks in 2025 were being evaluated against Rule of 60 expectations at the time of valuation. This is not yet a standard benchmark — it applies primarily to AI-native SaaS with a specific financial profile — but it is a real framework being used in practice.
Churn is the percentage of customers (logo churn) or revenue (revenue churn) lost in a given period. Monthly and annual churn have very different implications for LTV and NRR.
The relationship between monthly and annual churn is not linear. Annual churn rate ≈ 1 - (1 - monthly churn rate)^12.
| Monthly Churn | Approximate Annual Churn | Annual Customer Retention |
|---|---|---|
| 0.5% | ~6% | 94% |
| 1.0% | ~11.4% | 88.6% |
| 1.5% | ~16.5% | 83.5% |
| 2.0% | ~21.5% | 78.5% |
| 3.0% | ~30.6% | 69.4% |
| 5.0% | ~46% | 54% |
At 3% monthly churn, you are losing roughly 30% of your customer base every year. That is not a SaaS business — it is a revolving door. Many early-stage founders do not realize how small monthly churn numbers compound into enormous annual attrition.
| ACV Segment | Best-in-Class Annual Logo Churn | Acceptable | Concerning |
|---|---|---|---|
| SMB (< $5k) | < 10% | 10-20% | > 25% |
| Mid-market ($5k-$50k) | < 7% | 7-12% | > 15% |
| Enterprise ($50k-$200k) | < 5% | 5-8% | > 10% |
| Strategic Enterprise ($200k+) | < 3% | 3-6% | > 8% |
| ACV Segment | Best-in-Class Annual Revenue Churn | Acceptable | Concerning |
|---|---|---|---|
| SMB (< $5k) | < 8% | 8-15% | > 20% |
| Mid-market ($5k-$50k) | < 5% | 5-10% | > 12% |
| Enterprise ($50k-$200k) | < 3% | 3-6% | > 8% |
| Strategic Enterprise ($200k+) | < 2% | 2-4% | > 6% |
When revenue churn is lower than logo churn, it means your larger customers are staying while smaller ones are leaving. This is usually acceptable and sometimes desirable — if you are intentionally moving upmarket, losing small accounts while retaining large ones reflects the strategic direction.
When revenue churn is higher than logo churn, you have a problem. It means your largest customers are churning at a higher rate than small ones, which destroys both revenue and the statistical foundation of your LTV calculations. This pattern requires immediate investigation.
Gross margin is the percentage of revenue remaining after cost of goods sold (COGS). For SaaS, COGS includes hosting, infrastructure, customer support, and professional services directly tied to revenue generation.
| Product Type | Best-in-Class GM | Typical Range | Below Average |
|---|---|---|---|
| Pure SaaS (no services) | > 80% | 70-80% | < 65% |
| SaaS + professional services | > 70% | 60-72% | < 55% |
| Usage-based SaaS (API, consumption) | > 70% | 60-75% | < 55% |
| AI-native SaaS (LLM API costs) | > 65% | 55-70% | < 50% |
| Infrastructure / DevOps SaaS | > 75% | 65-78% | < 60% |
| Vertical SaaS (hardware component) | > 60% | 50-65% | < 45% |
AI-native SaaS products built on top of third-party LLM APIs have structurally higher COGS than traditional SaaS because the cost of inference is embedded in delivery. Every customer interaction that invokes a model call has a marginal cost attached to it.
In 2023 and early 2024, this was a serious concern — companies building on GPT-4 were seeing COGS that compressed gross margins to 40-50%. The rapid decline in inference costs through 2024 and 2025 has changed this picture substantially. Companies that were achieving 50% gross margins in 2023 are reporting 65-70% margins in 2025 on the same product, primarily because the cost of model inference has fallen by an order of magnitude.
The 2026 benchmark for AI-native SaaS gross margins of 55-70% reflects this improved cost environment. Companies building efficient AI products today should be targeting the higher end of this range. Companies still at 50% or below need to either improve their inference efficiency or build in more margin through feature expansion.
These two metrics became central to fundraising conversations after the 2022 correction, and they remain primary filters for most growth-stage investors in 2026.
The Magic Number measures how much net new ARR you generate for every dollar of sales and marketing spend in the prior quarter.
Magic Number = (Net New ARR in Quarter N) / (S&M Spend in Quarter N-1)
| Magic Number | Interpretation |
|---|---|
| > 1.5 | Exceptional; accelerate sales investment |
| 1.0 - 1.5 | Strong; healthy to invest in growth |
| 0.75 - 1.0 | Good; worth investing at current level |
| 0.5 - 0.75 | Acceptable; proceed cautiously |
| < 0.5 | Concerning; fix GTM efficiency before scaling |
| Stage | Best-in-Class | Good | Acceptable | Concerning |
|---|---|---|---|---|
| Seed | > 1.5 | 1.0-1.5 | 0.75-1.0 | < 0.5 |
| Series A | > 1.5 | 1.0-1.5 | 0.75-1.0 | < 0.5 |
| Series B | > 1.0 | 0.75-1.0 | 0.5-0.75 | < 0.4 |
| Series C+ | > 0.75 | 0.6-0.75 | 0.5-0.6 | < 0.4 |
Burn Multiple is the most important efficiency metric added to the standard SaaS toolkit post-2022. It was popularized by Benchmark partner Bill Gurley and measures how much capital is being consumed for each dollar of net new ARR generated.
Burn Multiple = Net Cash Burned / Net New ARR
| Burn Multiple | Interpretation |
|---|---|
| < 1x | Exceptional; every dollar burns generates >$1 ARR |
| 1x - 1.5x | Strong; efficient growth |
| 1.5x - 2x | Acceptable for early stage |
| 2x - 3x | Borderline; scrutinized by investors |
| > 3x | High concern; unsustainable capital consumption |
| Stage | Best-in-Class | Good | Acceptable | Concerning |
|---|---|---|---|---|
| Seed | < 2x | 2-3x | 3-5x | > 5x |
| Series A | < 1.5x | 1.5-2x | 2-3x | > 3x |
| Series B | < 1.25x | 1.25-2x | 2-2.5x | > 2.5x |
| Series C+ | < 1x | 1-1.5x | 1.5-2x | > 2x |
The burn multiple has become the single most scrutinized metric in late-stage venture capital since 2022. A company burning $20M per quarter to generate $5M in net new ARR has a 4x burn multiple. That company will struggle to raise a growth round unless it has an extremely compelling story about why the efficiency will improve dramatically and quickly.
Before 2022, growth rate was the dominant metric for valuation. Companies growing 150% YoY could raise at 50x revenue regardless of profitability. The correction changed this permanently.
In the current environment, growth rate and efficiency are both required. The typical Series B investor in 2026 uses a mental model roughly equivalent to: "I want growth rate + profitability margin above 40, and I want burn multiple below 2x." The combination of these two filters eliminates a large percentage of companies that would have been fundable in 2021.
AI-native SaaS — products built from the ground up with AI as a core capability rather than a feature addition — is producing a new set of benchmark deviations that the standard framework does not fully capture.
ARR growth velocity: The top quartile of AI-native SaaS companies at Series A are growing at rates that significantly exceed the traditional T2D3 framework. Several AI infrastructure and AI application companies reached $10M ARR in under 12 months from product launch in 2024-2025. The speed of AI product adoption is driving growth curves that did not exist in traditional SaaS.
Product-led growth efficiency: AI-native products often have stronger PLG fundamentals than traditional SaaS because the product's value is demonstrable immediately (you can generate output in seconds) and the product becomes more valuable with use. This compresses CAC for companies that have built strong self-serve motions.
Expansion revenue: Usage-based pricing models common in AI SaaS (where customers pay per API call, per generation, or per output) create natural expansion mechanics. As customers use the product more, they spend more, without any upsell motion required. This drives NRR above 130% for top-quartile AI SaaS companies with usage-based pricing.
Gross margin compression: As discussed, LLM inference costs structurally compress gross margins for AI-native products. Best-in-class AI SaaS at 70% gross margin is competing against best-in-class traditional SaaS at 80%+ gross margin. This matters for long-term valuation multiples.
Churn risk from model commoditization: AI products built primarily on one provider's model face churn risk if a competitor model becomes substantially better or cheaper. Customers who chose a product for its AI quality may switch when a better model is available through a different provider. Companies without strong switching costs beyond the model quality are more vulnerable to this pattern.
Burn multiple during model experimentation: Early-stage AI companies often have elevated burn multiples during the model selection and fine-tuning phase. This is capital-intensive work that does not directly produce ARR. Investors in AI SaaS have shown more tolerance for elevated burn multiples at early stages compared to traditional SaaS, but that tolerance decreases sharply at Series B and beyond.
| Metric | Traditional SaaS Benchmark | AI-Native SaaS Adjustment |
|---|---|---|
| Gross Margin (Series A) | > 70% | > 60% acceptable if trajectory is improving |
| NRR (Series A) | > 110% | > 115% expected (usage expansion advantage) |
| ARR Growth (Seed to A) | 150-300% | 200-500% realistic; top quartile is higher |
| CAC Payback (Series A) | < 18 months | < 12 months for PLG-led; < 24 for enterprise |
| Burn Multiple (Series A) | < 2x | < 2.5x accepted if product is capital-intensive build |
| Rule of 40 (Series C) | > 40 | Rule of 60 is the emerging target for top performers |
Benchmarks are reference points, not targets. Here is how I recommend using them in practice.
This seems obvious, but most founders who ask me about benchmarks do not have accurate calculations of their own metrics. Before benchmarking anything, make sure you can answer:
If you cannot answer these questions with confidence, fixing the measurement problem is more important than benchmarking.
Stage and ACV segment are your two primary peer group filters. A seed-stage SMB SaaS should not be benchmarking against Series C enterprise SaaS. Get the stage and segment right before comparing numbers.
A third dimension that is often overlooked: go-to-market motion. A product-led growth company has fundamentally different CAC and NRR dynamics than a field sales enterprise company, even at the same stage and ACV. If you are PLG, find PLG benchmarks. If you are enterprise field sales, find enterprise benchmarks.
Use this decision tree:
Not all metrics are weighted equally by investors at every stage. Here is how the weighting shifts:
| Stage | Primary Metric Focus | Secondary Metric Focus |
|---|---|---|
| Seed | ARR growth + founder story | CAC efficiency (early signal) |
| Series A | ARR growth + NRR | CAC payback + burn multiple |
| Series B | Rule of 40 + NRR | Gross margin + burn multiple |
| Series C+ | Rule of 40 + efficiency | Path to profitability |
| Metric | Seed Good | Series A Good | Series B Good | Series C+ Good |
|---|---|---|---|---|
| ARR YoY Growth | > 150% | > 100% | > 60% | > 40% |
| NRR | > 100% | > 110% | > 115% | > 120% |
| GRR | > 80% | > 87% | > 90% | > 92% |
| CAC Payback | < 18 mo | < 18 mo | < 18 mo | < 18 mo |
| LTV:CAC | > 3x | > 4x | > 5x | > 6x |
| Gross Margin | > 65% | > 70% | > 72% | > 75% |
| Burn Multiple | < 3x | < 2x | < 1.5x | < 1x |
| Magic Number | > 0.75 | > 1.0 | > 0.75 | > 0.6 |
| Rule of 40 | — | > 30 | > 35 | > 40 |
| Annual Logo Churn (mid-mkt) | < 15% | < 12% | < 10% | < 8% |
NRR. Here is why: at seed and Series A, ARR growth can come from an unsustainable burst of new customers who later churn. NRR tells you whether the customers you have acquired are staying and growing. An early-stage company with 115% NRR has fundamentally validated that its product has real, recurring value. An early-stage company with 85% NRR is sitting on a leaking bucket — it can grow through the top but it will struggle to scale efficiently.
Three things changed that are unlikely to revert. First, burn multiple is now a primary filter at every stage — a high burn multiple requires a compelling narrative rather than being accepted as table stakes for growth. Second, gross margin expectations have hardened; SaaS companies below 65% gross margin face valuation multiple compression regardless of growth rate. Third, the "growth above all else" narrative is gone. Growth rate without efficiency is no longer a story investors fund at premium valuations.
NRR above 100% means your existing customers generate more revenue over time, which means your business has a built-in growth engine independent of new customer acquisition. For investors, NRR above 110% signals product-market fit, healthy customer economics, and a business that gets more valuable as it scales. It also implies the company can grow efficiently even through periods when new logo acquisition slows — which is a significant risk mitigation factor.
For enterprise SaaS with ACV above $100k, 18-30 months is commonly accepted. The logic: enterprise customers churn at very low rates (2-5% annually) and frequently expand, so the long-term economics support a longer payback period. A company with 30-month CAC payback, 3% annual churn, and strong expansion revenue may have a better business than a company with 12-month payback and 20% annual churn.
Pre-seed benchmarks are projections, not measurements. The relevant question at pre-seed is not "what are my metrics?" but "what assumptions am I making about these metrics, and are those assumptions defensible?" Investors at pre-seed evaluate the founder's understanding of the unit economics model more than the actual numbers. Show that you understand what your CAC will be and why, what retention you expect and from what proxy evidence, and what your gross margin structure will look like.
Top-quartile SaaS companies at Series B and beyond typically show 40-60% of net new ARR coming from expansion (upsell, cross-sell, seat expansion). This is healthy because expansion revenue has near-zero CAC compared to new logo acquisition. If expansion is below 20% of net new ARR at Series B, the company may be leaving significant growth on the table or may not have built the customer success infrastructure to drive expansion systematically.
Usage-based SaaS requires different calculations because there is no fixed contract value. NRR for usage-based is calculated on cohort revenue over trailing 12 months versus the prior year cohort revenue. CAC payback is calculated using average monthly consumption rather than a fixed MRR figure. The key metrics to watch in usage-based SaaS are: cohort expansion rate (how much more do customers spend in month 12 vs. month 1), logo retention rate (are customers still active), and revenue concentration (what percentage of revenue comes from the top 10 customers).
AI companies in active model training or fine-tuning phases are generally given more tolerance on burn multiple — up to 3-4x at seed and early Series A. The expectation is that once the model development phase stabilizes and the commercial GTM takes over, burn multiple will improve substantially. However, this tolerance has limits. A Series B AI company still above 3x burn multiple on the basis of ongoing model costs will face investor scrutiny, particularly because the market has seen enough AI companies to have calibrated expectations about what the trajectory should look like.
Vertical SaaS typically has lower NRR (fewer expansion opportunities within a fixed customer use case), lower churn (deep workflow integration creates high switching costs), and higher CAC in absolute terms but lower CAC as a percentage of ACV because vertical SaaS tends to command higher prices. The Rule of 40 benchmark applies equally to vertical and horizontal SaaS, but the components will be weighted differently: vertical SaaS often reaches Rule of 40 through margin (high pricing, efficient GTM) rather than growth (smaller addressable markets limit the growth ceiling).
The IPO bar has tightened significantly. Based on companies that successfully went public or received favorable late-stage marks in 2025-2026, the approximate benchmarks for IPO readiness are: $100M+ ARR, ARR growth above 30% YoY, NRR above 115%, gross margin above 72%, Rule of 40 score above 40, and a credible path to GAAP profitability within 12-18 months of IPO. The Rule of 60 target for AI-native SaaS reflects how high the efficiency bar has been set for the most competitive public market entrants.
Learn how to calculate customer acquisition cost correctly — including every hidden cost most founders miss — with worked examples, benchmarks, and a complete tracking template.
MiniMax released M2.5 — a 230 billion parameter mixture-of-experts model with 80.2% on SWE-Bench Verified, matching Claude Opus 4.6 speed at $0.29/M input tokens, trained across 200,000 real-world coding environments.
OpenAI releases GPT-5.4 with major benchmark improvements, enhanced reasoning, and reduced hallucinations. Available now to ChatGPT Plus and API users.