I’ve been poring through public financial documents looking for ARR definitions and calculations. I could not find one pure-play usage company that defines ARR.
If these companies don’t define ARR, what metrics do they track?
In this post, I’ll walk you through the top financial metrics that usage-based companies track and disclose, why those metrics matter to management, and what Wall Street analysts care about when probing financial performance.
We’ll end with how these companies are communicating AI initiatives.
Financial Metrics Usage-based Companies Track
I researched the earnings call transcripts from Snowflake, Twilio, JFrog, and Fastly. I wanted to focus on pure-play usage companies or companies that have material usage-based revenue. Maybe, internally, they state an ARR but not the public.
Here are the common metrics tracked by these companies.
- Product or Core Revenue Growth — Always stated year-over-year and often tied to specific workload types or customer adoption patterns. When I use workload, I mean a distinct set of computing tasks, queries, transactions, or processes that a customer runs on the company’s infrastructure or service.
- Net Revenue Retention (NRR) or Dollar-Based Net Expansion (DBNER) — The expansion signal in a usage model, revealing how much more existing customers are consuming.
- Remaining Performance Obligations (RPO) — Forward visibility into contracted usage revenue, even if revenue recognition timing is flexible. It’s a requirement to disclose for public companies, and I do think it’s useful for private tech companies, especially in due diligence situations. Click RPO here for a post on this.
- Customer Counts by Segment — Especially large customers over a $100K or $1M ARR threshold. These companies liked to segment where they are showing traction.
- Gross Margin — Directly linked to efficiency improvements, network optimization, or mix shift to higher-margin revenue lines.
- Free Cash Flow and Operating Margins — Showing discipline and scalability even while investing in growth.
- AI/ML Workload Adoption — Now a headline KPI for many, either with quantified user counts or qualitative adoption signals. Often, only context was provided but some did provide quantification.
Here are some direct company quotes so you can see how they are used in context on these earnings calls.
Snowflake Quote
“Our net revenue retention rate was 128% for the trailing twelve months, highlighting our ability to expand existing customer relationships through increased consumption and new workloads.” — Mike Scarpelli, CFO
Twilio Quote
“Approximately 85% of inbound leads are now handled by our AI assistant, and those assisted leads are three times more likely to convert from free to paid.” — Khozema Shipchandler, CEO
JFrog Quote
“Enterprise Plus adoption continues to grow as customers bring AI model artifacts into our platform, reinforcing the role we play in securing and managing the AI software supply chain.” — Shlomi Ben Haim, CEO
Fastly Quote
“Security represented 47% of total revenue this quarter, up from 43% a year ago, showing our ability to cross-sell higher-margin solutions.” — Ronald Kisling, CFO
Key Financial Metrics Takeaways
- Usage companies consistently track a small set of financial and operational KPIs that reveal consumption, efficiency, and customer expansion.
- AI adoption is now part of that core metric set for many operators.
- Metrics are disclosed with context — tying numbers to product mix, customer segments, and workload trends.
Why These Financial Metrics Matter to Management
Management teams repeatedly use these KPIs to reinforce their growth narrative.
- Revenue Growth is the proof point for product-market fit and consumption health. Snowflake linked it directly to AI workload growth. Twilio tied software revenue growth to its shift away from declining communications volumes. Seems like Twilio’s SMS/email revenues are flat or declining?
- NRR / DBNER is the core of a usage pricing model. JFrog explained how security and AI workloads are lifting retention, while Snowflake cited enterprise expansions and new workloads.
- RPO reassures investors about pipeline health and deal quality. Multi-year contracts with large enterprises were a common driver. If RPO is important to you, move to multi-year contracts.
- Margins show scalability. Both Fastly and Twilio called out margin expansion from mix shift toward higher-value offerings. Margins by revenue stream is a “must master” financial concept for your tech business.
- AI Metrics serve as a forward-growth indicator. Snowflake quantified thousands of weekly AI/ML users, Twilio reported internal AI assistant adoption rates, and JFrog positioned AI as a platform stickiness driver.
Snowflake Quote
“Over 5,200 accounts are using our AI and machine learning capabilities weekly, embedding these workloads into their core operations.” — Srini Raghavan, SVP Product Management
Key Metrics Context Takeaways
- These KPIs aren’t just tracked. They are leveraged to tell a strategic growth story.
- Leaders use them to connect operational execution with financial outcomes. That’s the art of your FP&A team. Tying operations to the numbers.
- AI data is presented as either a quantitative KPI or a qualitative proof point, depending on maturity of their AI initiatives.
What Wall Street Analysts Want to Understand
The analyst Q&A section tells you what Wall Street is really watching. Analyst questions reveal the levers in the business and assumptions that they can use in their financial modeling.
- Consumption Patterns — Is growth seasonal, AI-driven, or tied to specific workloads?
- AI Impact — Is AI adoption incremental to revenue, and are AI workloads higher-margin or higher-consumption? Those are great questions!!!
- Retention Drivers — What’s fueling NRR above 120%? Which verticals or workloads are sticky?
- Margin Sustainability — Are gross margin gains structural or temporary?
- Pipeline Health — How does RPO split between short-term and long-term revenue?
- Guidance Conservatism — Is upside from AI adoption fully baked in, or could actuals beat forecasts? The great unknown of AI monetization.
Astute Analyst Questions
“Can you break down how much of your sequential consumption growth came from AI workloads versus traditional workloads?” — Analyst to Snowflake.
“Is the uplift in gross margins primarily from AI-related higher-margin workloads, or from network optimization efficiencies?” — Analyst to Fastly.
“Given your AI assistant adoption internally, how much of that efficiency gain can you translate into customer-facing monetization?” — Analyst to Twilio.
Key Analyst Takeaways
- Analysts push for clarity on what’s driving metrics and whether growth drivers are durable. We love durable growth!
- AI adoption and its direct impact on revenue and margins is a consistent point of questioning.
- SaaS and AI operators should expect investor focus on not just numbers, but the underlying behaviors and product dynamics driving them.
How Usage-based SaaS Leaders Communicate AI Initiatives
From Snowflake to Twilio to JFrog, and even Fastly to some extent, AI is no longer a side note in earnings calls. It’s integrated into product roadmaps, consumption models, and financial narratives.
But how they describe and measure AI initiatives varies. That offers lessons for founders looking to communicate their own AI stories.
Common Themes:
Tie AI to Existing Revenue Drivers
- Snowflake: AI workloads embedded in core consumption model, with named customer use cases (Siemens, Kraft Heinz, Samsung Ads).
- Twilio: AI adoption tied to funnel metrics (85% inbound lead handling, 3× upgrade likelihood) and customer automation (Cedar’s Kora agent automating 30% of inbound calls).
- JFrog: AI integrated into MLOps and security workflows, focusing on stickiness.
- Fastly: AI-driven edge optimizations boosting workload share.
Blend Quantitative and Qualitative Signals
- Metrics: User counts, workload percentages, conversion rates.
- Proof points: Named customers, AI-powered features, partnerships (NVIDIA, Hugging Face, ElevenLabs).
Position AI as a Long-Term Lever
- No one promises overnight AI transformation — AI is framed as a multi-year driver and moat builder.
Investor-focused Framing
- AI stories emphasize stickiness, upsell potential, and margin lift.
- Management addresses analyst questions around monetization and scalability.
Key AI Initiative Takeaways
- Investors respond to AI stories that are tied to measurable adoption and business outcomes.
- Founders should communicate AI impact in both usage terms (engagement, workload share) and business terms (retention, revenue, margins).
- When you read through earnings call transcripts for usage-based public software companies, a pattern emerges. They anchor the numbers in context, explain why they matter, and show how they’re tied to growth levers.
That’s it for today! If you enjoyed this post, please share in your Teams, Slack, or Discord channels.
I have worked in finance and accounting for 25+ years. I’ve been a SaaS CFO for 9+ years and began my career in the FP&A function. I hold an active Tennessee CPA license and earned my undergraduate degree from the University of Colorado at Boulder and MBA from the University of Iowa. I offer coaching, fractional CFO services, and SaaS finance courses.