Learn how to respond as AI changes the rules for operating metrics in software companies.
This piece is adapted from an Operators Guild Focus Session led by Geetha Rajan, exploring how AI is reshaping the operating assumptions behind modern software businesses. The discussion was grounded in her Pilots to Scale™ framework, which examines how organizations move from isolated AI pilots to measurable, organization-wide adoption — and how operating metrics must evolve as that transition happens.
Focus Sessions are small-group, member-only conversations where operators pressure-test decisions in flight and work through the realities that shape how companies are actually built and run.
If you want access to sessions like this — including recordings, working examples, and the community behind them — you can apply to join Operators Guild.
AI adoption is forcing operators to revisit assumptions that once felt settled.
For years, most software businesses followed a familiar economic pattern. Revenue scaled with seats, margins improved with scale, and growth rewarded predictability. Metrics like ARR, gross margin, and CAC payback offered a reliable way to understand performance and plan ahead.
AI disrupts that clarity.
Inference costs scale with usage. Value is often delivered dynamically rather than through static features. Customer behavior directly affects cost structures in ways traditional SaaS models were not designed to handle.
As discussed through the Pilots to Scale™ lens, the result isn’t confusion — it’s friction. The math still works, but it behaves differently, and operators are now responsible for interpreting systems that move in less linear ways.
Why AI breaks traditional SaaS unit economics
Traditional SaaS depends on high upfront development costs and low marginal costs at scale. Once a product is built, serving additional customers becomes cheaper over time.
AI changes that equation.
In many AI-powered products, usage introduces variable costs that rise alongside customer activity. Growth can increase operating expenses rather than smooth them out, creating tension between adoption, margins, and scale.
During the session, operators discussed practical responses, including revisiting how marginal cost behaves at different usage levels, modeling downside scenarios tied to customer behavior, and distinguishing between costs that can be optimized versus those that are structural.
How usage changes what “value” means
Usage-based pricing promises alignment. Customers pay for what they use. Companies grow when customers get more value.
In practice, usage introduces volatility. Revenue forecasting depends more on behavior than contracts. Engineering teams become responsible for both performance and cost efficiency. Go-to-market teams sell outcomes without guaranteeing consumption patterns.
A key takeaway from the discussion was the importance of clearly defining what customers are actually paying for — whether access, automation, insight, or labor replacement — and aligning internal metrics accordingly, a central theme within the Pilots to Scale™ framework.
Adapting metrics without overcorrecting
Not every legacy metric needs to be discarded. Retention still matters. Trust still compounds. Clear ownership still drives results.
What changes is how metrics are interpreted. Operators shared approaches such as pairing gross margin with usage-efficiency metrics, contextualizing ARR growth alongside cost-to-serve trends, and tracking customer outcomes rather than activity alone.
The goal, as emphasized in Pilots to Scale™, is not to abandon rigor, but to apply it where it matters most.
The operator challenge ahead
AI adoption reshapes more than dashboards. It changes workflows, expectations, and accountability. Friction emerges when teams feel the rules are changing without explanation.
Strong operators focus less on perfect forecasts and more on understanding how their systems behave under different conditions. Progress looks less like steady curves and more like disciplined course correction.
These are exactly the kinds of conversations happening inside Operators Guild every day.
If you want access to small-group discussions where senior operators compare notes, pressure-test assumptions, and work through how AI is changing real operating decisions, consider becoming a member.