We authored this piece after speaking with finance leaders, operators, and investors across the SaaS ecosystem who are all asking similar questions about AI:
How do we know whether AI is actually improving the business?
Is it driving incremental revenue? Is it changing the cost structure?
Are AI-enabled products scaling efficiently?
AI is rapidly becoming table stakes in SaaS products and operations through copilots, embedded models, usage-based pricing, and increasingly autonomous workflows. Yet many teams are discovering that while AI adoption is accelerating, their ability to measure its impact is lagging.
The issue isn’t a lack of data. It’s that AI changes what needs to be measured. Traditional SaaS metrics weren’t designed to isolate AI’s impact on revenue, costs, or productivity. Without updating those definitions, even sophisticated dashboards can give a misleading picture.
This post introduces the problem, but it doesn’t attempt to answer those questions fully. We dive deeper in our newly released paper, Measuring the Impact of AI: Getting the Right Metrics in Place, which lays out a practical measurement framework and the specific revenue and cost metrics finance and ops leaders need to focus on to evaluate AI impact rigorously.
Key takeaways:
• If you do not change your metrics, you cannot evaluate AI impact. AI adoption requires more granular data, not just new tools.
• AI revenue needs to be separated from the core business. In usage-based models, ARR alone will not reflect what is really happening.
• Different types of AI products behave differently. Tracking traditional ML features, AI add-ons, and agentic products separately shows what is actually working.
• AI introduces new, material cost drivers. Model costs, infrastructure, and AI-focused R&D need to be tracked explicitly and kept separate from general IT spend.
• Once AI-specific revenue and cost metrics are in place, standard productivity metrics like revenue per FTE or hosting costs as a percent of revenue become actionable again.
A Practical Framework for AI-Aware Metrics
At OPEXEngine, we work with SaaS operators and investors every day to benchmark performance and guide resource allocation decisions. Based on that work, we’ve developed a practical framework for putting the right AI-specific revenue and cost metrics in place—without overcomplicating reporting or losing comparability over time.
The full framework covers:
• How to structure AI revenue so signal isn’t drowned out by noise
• How to make AI costs visible and comparable
• How to restore the usefulness of standard SaaS productivity metrics in an AI-driven world
AI may be new, but the fundamentals of operational excellence still apply. The companies that win won’t just be the ones adopting AI fastest. They’ll be the ones measuring its impact most rigorously.
Download the full guide to see:
• Detailed metric definitions
• Revenue and cost breakout examples
• How to apply AI-aware metrics to benchmarking and resource allocation decisions
