Understanding SaaS and Cloud Economics

ARR Growth and Market Valuation

SailPoint (SAIL) reported fiscal second-quarter 2027 annual recurring revenue of $1.231 billion, up 25% year over year, with SaaS ARR increasing 36% to $847 million and net-new SaaS ARR accounting for 97% of total net-new ARR, according to the company’s Q2 FY2027 results. Yet as of September 12, 2026, SAIL traded roughly 20% below its $22 IPO price while identity peer Okta (OKTA) had gained 102% year to date, per 247WallSt’s comparison. ARR growth alone does not account for the difference.

Growth determines the size of the business, but the market evaluates how that growth is generated: how much revenue comes from existing customers, how much gross profit remains after delivery costs, and how efficiently new revenue is acquired. ARR, NRR, GRR, gross margin, CAC payback, and Rule of 40 are the key metrics that link product and operational decisions to valuation.

Key Takeaways

Median B2B SaaS net revenue retention was 102% in the 2025 benchmark year, with usage-based models at 108% and seat-based models at 98%, per Aleph and Benchmarkit’s 2026 benchmark report.
Median software gross margin held at 80% (76% blended with services), but usage-only pricing models run at 62% because consumption revenue involves higher infrastructure and compute costs.
The median Rule of 40 score rose to 25% in 2025 from 15% the year before, driven almost entirely by cost reductions rather than faster growth, which slowed to 20%.
Median CAC payback improved to 16 months from 18, and the median Magic Number exceeded 1.0 for the first time in years at 1.37.
Median SaaS EV/EBITDA multiples for PE deals dropped to 11.7x from 20.4x in H1 2026, aligning more closely with non-SaaS multiples as the decade-long SaaS premium narrowed.

Net Revenue Retention and Customer Expansion

Net revenue retention measures how much recurring revenue remains from an existing customer group after accounting for expansion, contraction, and churn. Gross revenue retention measures the same but excludes expansion, so it maxes out at 100%. The difference between the two reflects expansion revenue. In the 2026 Aleph and Benchmarkit benchmark report, based on 342 B2B SaaS and AI-native companies’ full-year 2025 results, median NRR was 102% and median GRR was 84%. That 18-point difference means expansion is compensating for actual churn.

The split by pricing model is the clearest indicator. Usage-based companies had median NRR of 108%, while seat-based companies were at 98%, below the break-even point. Seat-based models face pressure as AI-driven headcount efficiency allows customers to consolidate roles or deploy agents that reduce their seat count. A seat-based company at 100% NRR is doing better than its peers; a usage-based company at 100% is doing worse.

Gross Margin as Cost Efficiency

Gross margin reflects how cloud architecture affects the income statement. Compute, storage, egress, support, and third-party services determine how much of each revenue dollar remains after delivery. The 2026 Aleph and Benchmarkit data shows median software gross margin at 80%, with top-quartile companies exceeding 86%, and blended total-revenue margin at 76% once services are included. Total gross margin below 70% usually indicates a cost-structure issue rather than a pricing problem.

Usage-only pricing had median gross margin of 62%, well below the 76% to 84% range of subscription models, because consumption revenue involves higher infrastructure and compute costs, especially during the ramp-up period before economies of scale reduce per-unit costs. This trade-off explains usage pricing’s other benefits: it delivers the highest NRR (108%) and ARR per employee ($291,000), but at lower margin.

Margin also improves with size. Software gross margin increases from 72% at under $5M ARR to 86% in the $50M to $100M range, a 14-point gain through operational use. At that scale, the range tightens (84% to 90%), meaning execution quality rather than structural factors determines the margin.

The AI impact is more complex than headlines suggest. Software gross margin stayed steady at 79% to 81% over four years, and the report could not confirm AI-driven margin pressure with the available data. Separately, ICONIQ’s 2026 State of AI snapshot, surveying about 300 executives building AI products, projects average AI gross margin improving to roughly 52% in 2026 as routing and caching mature, with most workloads handled by smaller models and only complex tasks sent to frontier models. Builders now use about 3.1 model providers on average, up from 2.8 six months earlier. A legacy SaaS company with AI features that summarize content keeps 80% margin because AI usage is minimal, while a company where the model handles every interaction incurs inference costs on each one. Both call themselves AI-powered, and benchmark medians combine their results.

The Rule of 40 as Composite Indicator

The Rule of 40 adds revenue growth rate to profit margin, with 40% as the target. The 2026 benchmark found the median B2B SaaS score increased to 25% in 2025 from 15% the year before, the largest single-year rise in five years of data, while top-quartile companies exceeded 43%.

The improvement came almost entirely from profitability, not faster growth. In that year, the median company cut R&D spending by 8 points of revenue (from 35% to 27%), reduced S&M by 2 points, and trimmed G&A, while growth slowed to 20%. The Rule of 40 improved because companies spent less, even as growth slowed. Whether this efficiency is structural or a one-time adjustment remains uncertain, as the report notes.

The metric becomes meaningful around $20M ARR and above; below that, growth rates and margin swings are volatile enough that the combined score fluctuates accordingly. For early-stage companies, growth rate and CAC payback provide clearer signals.

Pricing model also affects the metric. Subscription-plus-usage hybrid companies had the highest 75th percentile at 43%, just above the threshold, because they combine predictable base revenue with consumption-driven expansion. Non-seat subscription companies trailed with an 18% median, often because they are in an active market-expansion phase where investment has not yet translated into efficiency.

Metric	Median (FY2025 data)	Top quartile	Source
Software gross margin	80%	86%+	Aleph and Benchmarkit 2026
Net revenue retention	102%	110%	Aleph and Benchmarkit 2026
Gross revenue retention	84%	See source	Aleph and Benchmarkit 2026
CAC payback	16 months	6 months or fewer	Aleph and Benchmarkit 2026
Rule of 40	25%	43%+	Aleph and Benchmarkit 2026

Financial chart showing SaaS revenue and margin trends — SaaS financial reviews now connect revenue metrics to cloud usage and cost-to-serve data instead of treating them as separate analyses.

Sales Efficiency and Cost of Growth

CAC payback measures how long it takes to recover acquisition cost through gross profit. The 2026 benchmark reports the median at 16 months, improved from 18 months in 2024, with top-quartile companies recovering CAC in six months or less and the bottom quartile at 24 months or more. The longest payback in the sample was 48 months, meaning four years to recover acquisition cost before generating margin.

Two factors significantly affect this number. Payback should use gross-margin-adjusted revenue, not raw ARR, because ignoring delivery costs overstates performance for low-margin businesses. Also, S&M spend should be lagged, since the spend that closed this quarter’s new ARR mostly occurred in prior periods. Both are common reasons the metric is reported too optimistically.

The Magic Number, which compares incremental recurring revenue to prior-period sales and marketing spend, exceeded the critical 1.0 threshold for the first time in years at a median of 1.37. The blended CAC ratio dropped to $1.30 of S&M per dollar of new ARR, down 7% year over year, and the new-name CAC ratio improved about 19% to $1.63. Efficiency improved because companies cut spending and targeted better, not because they increased spending.

Segment differences matter more than the blended average. Sub-$5K ACV customers recover CAC in a median 11 months through low-touch digital acquisition, while $50K to $100K enterprise deals take 22 months. Horizontal B2B SaaS recovers CAC faster at the median (14 months versus 18 for vertical), but vertical SaaS earns it back through retention, posting a higher lifetime-value-to-CAC ratio (5.6x versus 4.1x). Comparing an enterprise sales motion to the all-in 16-month median can misrepresent its performance.

Architecture and Unit Economics

Tenant architecture is both a financial and technical decision. Multi-tenant systems share infrastructure across customers, which can improve resource use, simplify upgrades, and spread fixed platform costs over a larger base. Single-tenant systems isolate each customer, supporting compliance, data separation, and customization, but usually increase costs and operational complexity through duplicated infrastructure, separate upgrades, and customer-specific troubleshooting.

The common mistake is pricing a single-tenant service like a multi-tenant product. Dedicated environments require a higher base price because the vendor loses shared-resource efficiencies. When dedicated deployment is sold as a discount rather than a premium feature, margin damage may not appear until support and infrastructure costs accumulate. The opposite mistake is forcing every regulated customer into the shared model and losing high-value accounts that require isolation, data residency, or audit controls.

For cloud-storage products, cost drivers include compute, storage, egress, support, and monitoring, each behaving differently. Compute costs can be reduced through scheduling, caching, batching, and autoscaling. Storage depends on data volume, redundancy, tiering, and retention policies. Egress is a major margin drain: a customer storing large volumes but rarely retrieving files may look profitable under a capacity-based plan, while a customer frequently restoring, syncing, and exporting files can generate much higher costs under the same plan. The practical approach is to model cost per useful workflow, since file restore, preview, sync events, and compliance exports each have different costs and values.

Market Dynamics and AI Impact

The valuation environment shifted significantly in 2026. Forvis Mazars and PitchBook’s H1 2026 State of SaaS report found median SaaS enterprise value/EBITDA multiples for PE deals dropped to 11.7x from 20.4x, nearly matching non-SaaS multiples as AI disruption risk and higher capital costs reduced the long-standing SaaS premium. Only 23 SaaS companies went public in H1 2026, and PE deal volume fell to $58.8 billion across 702 deals. Industry success now depends on execution and fundamentals rather than momentum, and companies that exclude AI-related costs from core operations face increasing pressure to show underlying cash-flow strength.

A company can meet the Rule of 40 while having weak unit economics in one product line, or report strong NRR because a few large customers expanded while smaller ones churned. A clearer assessment combines the composite score with ARR mix, NRR by cohort, and gross margin by segment. If high NRR comes from usage growth that also improves gross margin, expansion is high quality. If it comes from usage that increases egress, replication, or support costs, revenue quality may be lower than the headline suggests.

AI has mixed effects. ICONIQ’s 2026 State of AI snapshot shows AI gross margins improving as routing and caching mature, and 37% of companies plan to change their AI pricing model in the next 12 months, driven by customer demand, competitive pressure, and margin concerns. Hybrid models with a modest platform fee plus usage charges and safeguards like annual commitments are becoming the practical standard. The risk is a company adding costly AI features without monetization, where the feature helps demos but harms the income statement.

Data center server racks representing cloud infrastructure costs — Infrastructure cost connects architecture choices to SaaS gross margin.

Strategic Implications for Investors and Founders

The SailPoint example explains why no single metric determines valuation. SailPoint grew ARR 25% to $1.231 billion and increased SaaS ARR 36%, but it also reported a GAAP operating loss of $59 million compared to $41 million a year earlier, while maintaining 20% adjusted operating margin, per its Q2 FY2027 results. Its AI-driven ARR exceeded $70 million, and existing customers adopting the AI-driven solution increased annual spending by more than 60%. The market still valued the stock below Okta, indicating investors focus on the path to sustainable cash flow, not just growth rate.

For founders and CFOs, the approach is to define the unit clearly, coordinate finance and engineering definitions, and track cloud COGS at the decision level. ARR should match billing, NRR and GRR should track cohorts, gross margin should use cost categories engineering can influence, and CAC payback should use gross profit rather than revenue. For a cloud-storage product, the weekly dashboard should include ARR changes by new, expansion, contraction, and churn; retention by cohort; cloud COGS by storage, compute, transfer, and support; and sales efficiency by channel.

For investors, examine every metric in detail. Companywide gross margin can conceal weak enterprise deployments. Companywide CAC payback can hide one channel that works and another that loses money. Companywide NRR can mask a small set of large expanding accounts while the broader base weakens. Companies that succeed in 2026 will convert recurring revenue into retained revenue, retained revenue into high-quality expansion, and expansion into gross profit.

I expect the median B2B SaaS Rule of 40 score to reach 30% or higher in the next Aleph and Benchmarkit benchmark report, because the 2025 improvement came from a structural cost reset in R&D and S&M that companies are unlikely to reverse while growth remains near 20%, and because usage-based and hybrid pricing models are increasing retention faster than seat-based models. Separately, I expect SailPoint (SAIL) to report fiscal 2027 ARR above $1.375 billion by its Q4 FY2027 results, based on the company’s guidance of $1.375 billion to $1.385 billion and its AI-driven ARR growing faster than the overall base.

More in-depth coverage from this blog on closely related topics:

Sources and References

Sources cited while researching and writing this article: