Comprehensive market data showing how granular usage event tracking supports accurate billing, auditable finance workflows, and strategic pricing for SaaS and AI companies
Usage-linked monetization is now a material part of SaaS revenue design. In SBI/Price Intelligently's 2025 survey of 321 SaaS operators and executives, 21.2% reported a usage-based pricing structure, while current market research projects the global SaaS market to grow from $255.79 billion in 2026 to $563.66 billion by 2034, a 10.38% CAGR. As consumption pricing and AI-linked costs expand, the ability to meter, store, and act on raw usage events becomes foundational to accurate billing and adaptable revenue design. Orb's raw usage event architecture retains granular usage data, while Orb Accuracy supports backfills, backdated price changes, and historical usage correction workflows.
Key takeaways
- Usage-based pricing is material, but revenue design is increasingly hybrid: 21.2% of respondents in SBI/Price Intelligently's 2025 SaaS survey used a usage-based pricing structure, while more than 80% of respondents used some form of seat-based pricing metric and only 8% relied on seats as their sole value metric
- Variable revenue is more visible at larger scale: In High Alpha's 2025 benchmark, 40% of SaaS companies above $50 million ARR included consumption- or outcome-based revenue in total ARR
- AI is accelerating consumption-linked billing: 36% of AI tools in BetterCloud's surveyed environments use only usage-based or token-consumption billing
- A data execution gap persists: More than 40% of SMB respondents were not using financial and usage-data analyses to guide pricing decisions, while 23% of respondents with a usage component directed customers to Sales or Customer Success to discuss overages
- Financial impact is measurable, but benchmarks need careful scope: McKinsey reports 5% to 10% margin growth for retail and consumer clients using dynamic pricing, while 61% of IT leaders in Zylo's 2026 research said unplanned SaaS cost increases forced them to cut projects
- Event and consumption volumes are already enormous: Mixpanel's 2026 benchmark corpus spans over 3.7 trillion tracked events, while a16z and OpenRouter analyzed more than 100 trillion tokens of real-world LLM usage
The state of usage-based billing: market growth and adoption
The shift toward consumption-linked pricing reflects a broader change in how software companies align revenue with the value customers receive. These statistics illustrate the scale, economics, and execution requirements behind that transformation.
1. 21.2% of survey respondents used a usage-based pricing structure
In SBI/Price Intelligently's 2025 survey of 321 SaaS operators and executives, 21.2% reported a usage-based pricing structure. The figure shows that usage-based pricing is already a meaningful model, but it is not a market-wide default, which makes flexible support for subscription, usage, and hybrid structures especially important.
2. 40% of SaaS companies above $50 million ARR included consumption- or outcome-based revenue in ARR
High Alpha's 2025 benchmark found that 40% of companies above $50 million ARR included consumption- or outcome-based revenue in total ARR. The share was lower among smaller ARR cohorts, indicating that variable monetization is particularly visible in larger SaaS businesses with more mature pricing portfolios.
3. Consumption-pricing respondents reported 43% YoY growth versus 34% for subscription-pricing respondents
High Alpha's pricing-model benchmark found that companies using consumption-based pricing reported 43% year-over-year growth, compared with 34% for subscription pricing, 40% for hybrid pricing, and 65% for outcome-based pricing. These are observational cohort comparisons, not evidence that the pricing model itself caused the growth.
4. Global SaaS market projected to reach $563.66 billion by 2034
Current market research estimates the global SaaS market at $255.79 billion in 2026 and $563.66 billion by 2034, representing a 10.38% CAGR over the forecast period. Growth at this scale increases the value of billing infrastructure that can support changing price structures without sacrificing metering precision.
5. More than 40% of SMB respondents were not using financial and usage-data analysis to guide pricing decisions
SBI reports that more than 40% of SMB respondents were not using financial and usage-data analyses to guide pricing decisions. That execution gap is directly relevant to modern revenue design: collecting usage is not enough if teams cannot reliably connect it to pricing, customer behavior, and financial outcomes.
AI and the acceleration of usage-based billing
AI applications present billing challenges that traditional subscription models do not always address cleanly. Token consumption, API calls, agent runs, outcomes, and compute usage can vary substantially between customers and sessions, making precise metering and flexible value metrics increasingly important.
6. 36% of AI tools in BetterCloud's surveyed environments use only usage-based or token-consumption billing
BetterCloud's 2026 State of SaaS research, based on 525 IT and security professionals, reports that 36% of AI tools in surveyed environments are billed only on usage or token consumption. This is a survey-based benchmark rather than a census of all AI applications, but it demonstrates how prominent consumption-linked monetization has become in enterprise AI portfolios.
7. Organizations deploy an average of 27 AI-powered SaaS applications
With 27 AI-powered SaaS applications per organization on average in BetterCloud's study, metering requirements multiply. Each application can expose different value units and consumption patterns, reinforcing the need for billing systems that can adapt to varied usage models.
8. AI-powered applications account for roughly 22% of current SaaS portfolios
BetterCloud reports that AI-powered applications account for roughly 22% of the total SaaS portfolio in surveyed organizations. Orb's AI billing solution supports tokens, agent runs, outcomes, credits, and hybrid combinations with seats or platform fees. Orb's dimensional price groups support pricing across multiple usage dimensions, such as region, instance type, and environment, using a single pricing configuration for dimension combinations. Its metering product is built for high-volume event ingestion and usage spikes.
9. The average number of SaaS apps per organization increased 11% year over year
BetterCloud reports an 11% year-over-year increase in the average number of SaaS applications per organization. As software portfolios expand, the number of usage, spend, contract, and billing signals that organizations need to reconcile expands with them.
10. Organizations spend an average of $11,530 per employee on SaaS annually
Zylo's 2026 benchmark reports $11,530 in average annual SaaS spend per employee, with a median of $9,455. This level of per-employee spend increases pressure on buyers and sellers to make pricing understandable, attributable, and aligned with the value delivered.
The data layer: scale, measurement, and standardization
Modern revenue data is not only about whether telemetry is collected. It is about scale, measurement maturity, standardization, and whether event and cost data can be operationalized across product, pricing, finance, and customer workflows.
11. Fewer than 25% of surveyed companies use KPIs or dashboards to quantify internal AI impact
High Alpha reports that fewer than 25% of surveyed companies monitor KPIs or use analytics dashboards to quantify internal AI impact, while more than one-third rely on informal feedback. This is a measurement-maturity benchmark rather than a direct measure of billing accuracy, but it shows how far many organizations still have to go in turning activity data into operational decisions.
12. Mixpanel's 2026 benchmark corpus spans over 3.7 trillion events, 12,000+ companies, and 22 billion devices
Mixpanel's 2026 State of Digital Analytics benchmark covers over 3.7 trillion tracked events across more than 12,000 companies and 22 billion devices. These are product-analytics events, not necessarily billable usage events; the statistic illustrates the scale of behavioral telemetry rather than billing-platform throughput.
13. Mixpanel's B2B benchmark analyzed 577 billion events across 3.8 billion devices
Mixpanel's 2026 B2B benchmark analyzed 577 billion events across 3.8 billion devices. The event volume represented 27% year-over-year growth, while device volume increased 11%, showing how quickly the behavioral analytics footprint of business software can expand. These are not direct counts of billable usage events.
14. Mixpanel's AI benchmark analyzed approximately 290.8 billion AI events across 2.61 billion devices
Mixpanel's 2026 AI benchmark draws on approximately 290.8 billion AI events across 2.61 billion devices. These are behavioral analytics events rather than records of usage billed to customers. Their scale underscores the value of consistent event definitions when connecting product activity to monetization and cost analysis.
15. About 68% of large spenders in the survey ($100M+) use or are experimenting with FOCUS-formatted cost-and-usage data
Among large spenders in the survey ($100M+), the Linux Foundation reports that approximately 68% are using or experimenting with FOCUS-formatted cost-and-usage data, with another 18% planning to. Standardized cost and usage data is becoming part of the operating model for technology-value management.
Leveraging usage data for revenue optimization
Usage and event data can support expansion analysis, engagement measurement, overage management, and spend visibility when teams connect it to commercial workflows. The benchmarks below do not all establish causation from usage data, but they show the revenue and retention contexts in which granular data matters.
16. SaaS companies above $50 million ARR generate roughly 60% of new ARR from existing customers
High Alpha reports that companies above $50 million ARR generate roughly 60% of new ARR from existing customers. The benchmark does not say usage data causes that expansion, but it highlights why mature SaaS teams need reliable signals for product adoption, entitlements, consumption, and customer value as the installed base becomes a larger source of growth.
17. Hybrid-pricing respondents reported 105% average NRR
High Alpha reports 105% average NRR for hybrid-pricing respondents, compared with 102% for subscription pricing, 100% for outcome-based pricing, and 99% for consumption-based pricing. These are associations within the benchmark, not proof that hybrid pricing causes higher retention. They do show why billing systems need to handle fixed and variable components together rather than forcing teams into a single monetization model.
18. 98% of new users are inactive within two weeks in the median digital product
Amplitude reports that 98% of new users are inactive within two weeks for the median digital product in a benchmark covering more than 10,600 products across more than 2,600 companies. That steep drop-off makes event-level engagement analysis valuable for identifying which behaviors correlate with activation, retention, and eventual commercial value.
19. 23% of respondents with a usage component directed customers to Sales or Customer Success for overages
In SBI's survey, 23% of respondents with a usage component said customers were directed to Sales or Customer Success to discuss an overage. Orb's spend controls support real-time spend monitoring, spend-threshold alerts, credit-depletion actions, emails and system actions, and can flag accounts that may benefit from a new plan, giving teams a more automated way to operationalize usage and spend conditions.
20. 78% of IT leaders reported unexpected charges tied to AI features or consumption-based pricing
Zylo's 2026 research found that 78% of IT leaders experienced unexpected charges tied to AI features or consumption-based pricing in the prior year. This makes transparent metering and spend visibility a customer-experience requirement as well as a finance requirement.
Financial and operational impact of robust metering and billing systems
The business case for accurate metering extends beyond invoice correctness. Variable pricing and usage-driven software costs create measurable pressure on margins, budget predictability, and operating capacity.
21. McKinsey reports 5% to 10% margin growth from dynamic pricing for retail and consumer clients
McKinsey reports that its retail and consumer clients using dynamic pricing achieved 2% to 5% sales growth and 5% to 10% margin growth. This is a retail and consumer benchmark, not a SaaS billing or metering benchmark, so the percentage should not be generalized to software. The relevant revenue-design lesson is narrower: price experimentation depends on trustworthy data and disciplined testing.
22. 61% of IT leaders cut projects because of unplanned SaaS cost increases
Zylo reports that 61% of IT leaders cut projects because of unplanned SaaS cost increases. Consumption-linked pricing can improve alignment between cost and value, but without clear usage visibility and spend controls it can also create budget volatility that affects investment decisions.
23. Average annual SaaS spend reached $55.7 million even as application counts stayed roughly flat
Zylo reports $55.7 million in average annual SaaS spend, up 8% year over year, while average portfolios held at 305 applications and overall app counts declined 0.07%. The pattern shows that SaaS spend can rise even when portfolio size is roughly flat, increasing the importance of cost attribution and usage visibility. Orb's finance workflows maintain product and revenue data history and provide a documented trail with exportable summaries and finance-stack integrations for reconciliation and reporting.
24. a16z analyzed more than 100 trillion tokens of real-world LLM usage
The a16z and OpenRouter study analyzed more than 100 trillion tokens of real-world LLM usage. During the cited week, OpenRouter was processing more than 1 trillion tokens per day. Token throughput is not the same as a count of billable usage events, but workloads of this scale make metering, dimensional attribution, and historical usage data important infrastructure considerations.
Security, compliance, and risk considerations
Usage event data can carry security and compliance implications that affect billing architecture. Proper data handling protects both vendors and customers while enabling the traceability required for enterprise revenue operations.
25. 67% of IT professionals cite data loss or rogue AI agents as a greater organizational risk than lagging AI adoption
BetterCloud's accompanying data shows 67% of IT professionals pointing to data loss or rogue AI agents as the greater risk, versus 33% citing lagging AI adoption. Billing systems that process usage and revenue data therefore need strong access controls, governance, auditability, and security practices.
26. The global average cost of a data breach reached $4.99 million
IBM's 2026 study reports a $4.99 million global average cost per breach, with AI-enabled malicious breaches averaging $6 million and representing one in four malicious breaches. Orb maintains SOC 1 Type 2 and SOC 2 Type 2 reports and provides security, governance, and access-control capabilities for enterprise revenue infrastructure.
Choosing value metrics and hybrid pricing structures
AI monetization is pushing software companies to reconsider what they meter, what customers understand as value, and how fixed and variable components should work together.
27. 27 of 50 technical AI buyers preferred credits tied to recognizable work, versus 14 who preferred tokens
In a16z's survey of 50 technical AI buyers, 27 preferred credits tied to recognizable work while 14 preferred token-based pricing. The small sample should be interpreted accordingly, but the result is useful for revenue design: the best metered unit is not always the lowest-level infrastructure unit. Orb supports tokens, agent runs, outcomes, credits, and hybrid structures, allowing teams to price closer to the value unit their customers understand.
28. More than 80% of respondents used some form of seat-based pricing metric, but only 8% used seats alone
SBI states that more than 80% of respondents used some form of seat-based pricing metric, while only 8% used seats as their sole value metric. The report's chart separately shows 33% using seats as a primary value metric alongside other metrics and 37% using seats as a secondary metric. The source's stated benchmark supports hybrid and multi-metric revenue design rather than a simple seat-versus-usage binary. Orb's price modeling supports seat-based, hybrid, prepaid-credit, usage, contract commits, and dimensional structures in the same pricing foundation.
AI spend, FinOps, and revenue-design complexity
AI consumption is increasingly a finance and technology-value management problem, not only a product or engineering concern. These benchmarks show how quickly responsibility, budget growth, and model complexity are expanding.
29. 98% of State of FinOps respondents manage AI spend, up from 31% two years earlier
The Linux Foundation's 2026 State of FinOps survey of 1,192 respondents reports that 98% now manage AI spend, up from 31% two years earlier. AI usage has moved rapidly from a niche cost category into mainstream technology-value management.
30. 90% of FinOps practitioners are now asked to manage SaaS, up from 65% the previous year
The same State of FinOps research reports that 90% of practitioners are now asked to manage SaaS, up from 65% the previous year. Software usage and spend telemetry are increasingly part of the same operating discipline as cloud-cost management.
31. In a16z's 2025 survey, enterprise CIOs expected approximately 75% average LLM-budget growth over the following year
In a16z's survey of 100 CIOs across 15 industries, enterprise leaders expected approximately 75% average growth in LLM budgets over the following year. Rapid variable-cost growth raises the value of precise metering, cost attribution, and spend visibility.
32. The share of surveyed enterprise LLM spending funded by innovation budgets fell from 25% to 7%
The same a16z research found that the share of LLM spending funded from innovation budgets fell from 25% to 7%. AI consumption is moving into recurring operating budgets, where finance teams require more predictable measurement and control.
33. 37% of surveyed CIOs used five or more models, up from 29% the prior year
A16z reports that 37% of surveyed CIOs used five or more models, up from 29% the prior year. Multi-model workloads increase the number of dimensions, rate structures, providers, and cost signals that revenue and finance systems may need to attribute correctly.
34. AI-native application spend rose 108% year over year, including 393% growth at large enterprises
Zylo's 2026 SaaS Management Index reports that AI-native application spend rose 108% year over year, with large enterprises up 393%, while use of applications in its broader Artificial Intelligence category grew 181%. That pace of expansion makes consumption volatility and usage visibility central considerations for both software buyers and vendors.
Why raw usage event data architecture matters
The statistics above show that usage-linked pricing, event scale, AI cost growth, and FinOps responsibilities are all increasing the operational value of granular, auditable usage data. Persisting raw usage events rather than only aggregate totals gives teams more flexibility when they need to investigate history, change pricing, correct usage, or explain charges.
Retroactive adjustments become possible
When billing disputes arise, usage arrives late, or contracts change, systems that retain raw usage event logs can re-rate historical usage from source-level records. Orb's Accuracy supports backfilling missing events, backdating price changes, and recalculating invoices as underlying usage changes, reducing dependence on manual reconciliation against pre-aggregated totals.
Pricing changes can be tested against historical patterns
McKinsey's 5% to 10% margin-growth benchmark applies to retail and consumer dynamic pricing and should not be generalized as a SaaS result. Separately, pricing teams with historical event-level data can model proposed structures against actual product usage. Orb's Simulations uses real product usage data rather than assumptions to model the financial outcomes of a price change and compare scenarios side by side before anything ships.
Audit trails support compliance and customer trust
With the global average breach cost at $4.99 million and 67% of IT professionals in BetterCloud's analysis prioritizing data-loss or rogue-agent risk over lagging AI adoption, billing systems that handle usage data need strong security controls and clear lineage from raw usage events to invoices. Orb lets finance teams explore invoices down to the usage event, supporting billing investigation, reconciliation, and auditability while giving customers a clearer basis for understanding charges.
High-throughput scenarios require purpose-built infrastructure
BetterCloud reports that 36% of AI tools in its surveyed environments use only usage or token-consumption billing, while the a16z and OpenRouter study analyzed more than 100 trillion tokens and reported throughput above 1 trillion tokens per day during the cited week. Orb is built for this shape of workload. Its raw data layer can surface a log of every event at any time, its persistent store supports backdated price changes while maintaining an audit trail, and its additional aggregation capabilities can ingest billions of events per day.
Frequently asked questions
How does granular usage event data improve billing accuracy?
Raw usage event data can support transaction-level verification when events contain the timestamps and dimensional attributes needed to calculate charges. With Orb, finance teams can explore invoices down to the usage event, making billing questions easier to investigate and helping customers understand what generated a charge. If usage arrives late or needs correction, Orb Accuracy supports backfills and recalculation workflows.
What are the benefits of a raw data layer architecture for usage-based billing?
Persisting raw usage events rather than only aggregated totals makes retroactive price changes, backdated adjustments, historical analysis, and usage corrections more flexible. Orb's raw data layer retains event-level history, while Accuracy supports backfills and backdating. Orb's invoicing workflows preserve adjustment history and support voids, credit notes, backfills, retroactive price changes, and invoice recalculation where applicable.
How can companies use usage data to simulate and optimize pricing changes?
With comprehensive event logs, pricing teams can apply proposed rate structures to historical usage and inspect the projected impact before changes go live. Orb's Simulations uses real product usage data to model financial outcomes and compare scenarios side by side. SBI's finding that more than 40% of SMB respondents are not using financial and usage-data analyses to guide pricing decisions shows why reliable historical data and practical simulation tooling remain a meaningful execution advantage.
What role does metering and usage data play in financial compliance for SaaS businesses?
Usage data can be an important input to revenue accounting for usage-based arrangements, but invoice timing does not determine recognition, and consumption does not equal recognition in every arrangement. ASC 606 treatment depends on contract terms, performance obligations, transaction price, and when those obligations are satisfied. Orb's technical revenue-recognition guidance explains this distinction. For NetSuite users, Orb's NetSuite integration creates standard transaction objects, maps to existing NetSuite items, supplies exact line-level service periods, and feeds existing ARM workflows while NetSuite remains the accounting system of record.
Can usage event data enhance customer transparency and trust?
Yes, when the same pricing and usage context is exposed consistently to customers and finance teams. Orb's Experience Kit supports advanced real-time dashboards for usage across dimensions such as teams, locations, and metrics, while Orb's Revenue Recognition tooling lets finance explore invoices down to the underlying usage event. That shared visibility makes complex consumption charges easier to understand and billing questions easier to investigate.