SAP recently signaled a major shift in how it plans to charge for AI, moving away from subscription pricing in favor of usage-based monetization. The rationale is simple and hard to ignore: as AI automates work, the number of users no longer reflects the value delivered.
That message matters because SAP isn’t experimenting on the margins. When a company of that scale publicly rethinks pricing around consumption, it confirms what many software leaders already sense: usage-based monetization is moving from optional to inevitable.
But here’s the part that often gets overlooked. If you’re planning a move to consumption-based pricing, your first step isn’t pricing. It’s usage analytics.
Usage-Based Monetization Starts Long Before Pricing Changes
Across SaaS, enterprise software, and AI-driven products, teams often treat usage-based monetization as a commercial decision. They debate terminology, such as tokens versus credits, prepaid versus postpaid, or how to position hybrid pricing models.
Those conversations matter, but they are downstream.
Consumption-based licensing only works when organizations understand how their products are actually used. Without this foundation, pricing becomes guesswork, and guesswork leads to customer friction, revenue leakage, and internal mistrust.
Before finalizing your AI pricing strategy, you need clear answers to basic questions, including:
- What’s being used, and by whom?
- Which actions correlate with value, renewal, and expansion?
- What does normal usage look like across customers and segments?
- Where do usage spikes occur, and why?
Without this strategic insight, usage-based monetization fails to deliver value for both buyers and suppliers.
The Tree You Should be Planting Today
As the proverb goes: the best time to plant a tree was 20 years ago. The second-best time is today.
For software monetization, the modern version is straightforward:
If you think usage-based monetization is in your future, you should already be collecting usage analytics.
Even if your pricing model isn’t changing this year, the data you collect now becomes the baseline for future decisions. It gives product teams evidence to define value metrics. It gives finance teams confidence in forecasting. It gives customers transparency when consumption becomes part of the commercial relationship.
Delaying the implementation of software usage tracking until pricing decisions are made effectively means you’re pricing blind.
AI Makes Usage Analytics Non-Negotiable
AI doesn’t just increase usage. It changes usage patterns entirely.
AI workloads can scale rapidly and unpredictably, often driven by automation rather than human interaction. That volatility is why AI monetization conversations focus so heavily on AI tokens, capacity allocation, and consumption limits.
But none of those mechanisms work without accurate, real-time usage visibility, measured consistently, transparently, and in a way customers can trust.
You cannot monetize what you cannot measure.
Billing isn’t the Hard Part
Many usage-based monetization initiatives focus too narrowly on billing. Billing answers one question: how much was used?
Modern software monetization must also answer a more important one: what is the customer allowed to do right now?
That real-time entitlement decision is where consumption-based licensing either builds confidence or erodes it. Customers expect clarity. They want to see usage as it happens, understand limits, and avoid surprise costs. Internally, finance teams need auditable data, and product teams need the freedom to evolve packaging without reengineering applications.
This is why entitlement management and usage analytics must be designed together. Usage-based monetization isn’t just about invoices. It’s about governing access, consumption, and value delivery in real time.
Your Path to Usage-Based Monetization
For technology companies preparing for usage-based AI licensing, the most effective approach is incremental and evidence-driven.
- Establish a usage truth layer. Start collecting granular, accurate usage analytics now, even if pricing isn’t changing yet.
- Connect usage to entitlements. Define what customers are entitled to consume and under what conditions.
- Introduce consumption mechanics with guardrails that protect both customers and revenue.
- Iterate pricing and packaging based on real-world usage patterns, not assumptions.
This sequence allows teams to move toward usage-based monetization with confidence instead of urgency-driven compromises.
Alignment with Revenera
Revenera’s approach to software monetization is grounded in the understanding that usage-based pricing succeeds only when measurement, entitlement management, and governance work together.
Revenera enables real-time, auditable usage capture that supports consumption-based licensing models. Pricing and packaging can be adjusted through configurable rate tables without disrupting applications or customer workflows. Usage insights are designed to support both operational decision-making and customer transparency.
This foundation allows organizations to introduce usage-based monetization deliberately, adapt as markets evolve, and avoid the trust issues that arise when monetization outpaces measurement.
Download Your Free eBook: AI Monetization, Done Right
The Big Takeaway
SAP’s move toward usage-based monetization isn’t just news. It’s a signal.
If introducing AI tokens and consumption-based pricing is part of your roadmap, the most strategic move you can make today isn’t to finalize a pricing model.
It’s to start collecting the usage analytics you’ll need to price, package, and govern consumption effectively.
The best time to plant that tree was years ago. The second-best time is now.
If you’d like to discuss how to implement usage-based monetization, please contact us and a member of our team will be happy to talk.
Common Questions
How long should you collect usage data before changing your pricing model?
Ideally, collect at least a full business cycle of data, often six to twelve months, so seasonal peaks, renewal periods, and quarter-end spikes are captured. A shorter window can miss patterns that only appear at certain times of year, leading to thresholds or rates that don’t hold up. Collecting product usage data across different customer segments during that period also shows whether one pricing model can serve everyone or whether segments need different approaches.
What is a value metric, and how do you choose the right one?
A value metric is the unit that pricing is based on, such as transactions processed, documents generated, or workflows completed. The right metric grows as the customer gets more value, is easy for customers to understand and predict, and can be measured accurately. Usage analytics shows which actions correlate with renewal and expansion, making them strong candidates. Comparing options against common SaaS pricing models helps confirm the metric fits how your market already buys.
How can customers see their usage as it happens?
The most effective approach puts usage data directly in front of customers through dashboards in the product or customer portal, showing current consumption, remaining balances, trends over time, and alerts as limits approach. Embedded analytics lets customers check this themselves instead of waiting for invoices or account reviews. That transparency is what builds trust when consumption becomes part of the commercial relationship.
What typically causes usage spikes, and how should pricing handle them?
Common causes include AI agents and automated workflows running at scale, batch processing, seasonal or project-driven demand, onboarding new teams, and occasionally misconfigured integrations. Understanding the cause matters, because a spike from genuine value should be monetized, while one caused by an error may need a credit or a fix. Credit-based pricing gives customers a buffer to absorb legitimate bursts without renegotiating contracts, while alerts flag unexpected spikes before they turn into surprise bills.
