Usage-Based Pricing in 2025: Insights from Metronome’s Scott Woody
Plus: Updates from Trello, LaunchDarkly, and Jotform.
Welcome back to Good Better Best!
This week, I’m sharing notes from an awesome conversation with Scott Woody, the CEO of Metronome. Scott has some of the most nuanced thoughts I’ve heard on the future of usage-based pricing, and I learned a ton in our short discussion.
Before that, a quick announcement.
A couple months ago,
asked if I’d be interested in creating a course about pricing and packaging. A general rule I try to follow is when someone much smarter than me thinks I should do something — I try to listen.Fast forward to this week, and I’m pumped to announce Pricing for SaaS: Your Competitive Advantage — a cohort-based course designed to help operators iterate rapidly on pricing and packaging in the age of AI.
The course is hosted by the fine folks at Maven, and I’ll be teaching alongside my former consulting partner, Wilson Sadowski, who’s been in the trenches leading pricing at Google Cloud for the past four years.
If you’re a Product Manager, Product Marketer, or Founder looking to sharpen your pricing skills, this course is for you. Check it out and shoot me a note if there’s anything I can answer.
📰 Pricing News & Updates
Jotform updated its plans.
Trello added features.
LaunchDarkly added value to the free plan.
Tabnine introduced features.
Sourcegraph updated plan limits.
Today’s Post is brought to you by Metronome
Metronome is the world’s leading usage-based billing platform, trusted by companies like OpenAI, Anthropic, and Confluent to launch new products and pricing faster.
Everyone’s talking about usage-based pricing—but how many SaaS companies are actually using it, and why? Metronome took a deep dive into the market to see what’s actually going on in their new state of usage-based pricing report.
Key highlights:
📊 85% of surveyed software companies have adopted UBP
🏢 77% of the largest software companies use some level of UBP
🚀 64% of Forbes’ next billion-dollar startups offer UBP
Dive deeper into these trends and see how leading companies are approaching pricing in the full report.
Usage-based pricing is experiencing a renaissance, and few people understand this transformation better than Scott Woody, CEO of Metronome.
After leading monetization engineering at Dropbox for six and a half years, Woody founded Metronome to solve the complex challenges of usage-based billing. In our recent conversation, he took me through Metronome’s history, from infrastructure companies to the biggest players in AI, and beyond.
Scott shared fascinating insights about how AI is democratizing usage-based pricing and revolutionizing consumer subscription models. Given Metronome’s rapid growth, and the emergence of other usage-based billing engines, its clear that SaaS operators are making usage a focal point of their pricing strategy.
But here’s what everyone misses — the future of usage-based pricing could very well be charged “per seat.” Scott shares why below.
Infrastructure Origins
Metronome's journey began with infrastructure companies like Cockroach Labs and Cribble, who faced acute challenges with usage-based pricing.
"Infrastructure companies had the hardest internal pricing systems to manage," Woody explains. To put this in perspective, he notes that while Dropbox and Twilio had similar revenue, Twilio's billing team peaked at around 200 people compared to Dropbox's 50 – highlighting the complexity of usage-based systems.
But the challenges weren't just operational. Usage-based models required real-time monitoring.
Case in point, an early Segment customer accidentally ran up a $200,000 bill when their total bank account was only $5,000. Scenarios like this highlighted the need for better tools and visibility into usage patterns.
Enter Generative AI
The landscape shifted dramatically with the rise of AI.
"AI kind of makes every company usage-based," Woody observes. This transformation is happening at unprecedented speed, with even traditional software giants making rapid shifts. He points to Salesforce's Agent GPT announcement as a prime example: "Salesforce literally invented the seat model, and they're moving that fast."
As B2B players move towards charging for outcomes, the importance of streamlined operations and real-time monitoring will only increase.
The Consumer Evolution
Woody sees consumer services as the final frontier for usage-based pricing, and believes AI products are leading the way. He draws an interesting parallel with cell phone plans:
"Consumers actually kind of understand usage, they just hate it by default because of power companies... but cell phones, I think people get. They're like, 'Cool, yeah. Makes sense to me. It's like an amount, but if I use too much, I have to pay.'"
This hybrid approach is already evident in leading AI products.
ChatGPT Plus operates on a $20/month subscription model but incorporates usage limits and features that operate on a consumption basis. Subscribers get priority access and faster response times, but they're still bound by rate limits. Similarly, GPT-4 access comes with specific token limits that reset each day, effectively creating a usage-metered system within the subscription framework.
Perplexity offers another interesting example with their Pro tier at $20/month. While it's packaged as a flat subscription, it includes usage-based components like daily limits on GPT-4 and Claude queries, alongside unlimited GPT-3.5 usage. This creates a tiered consumption model that feels predictable to consumers while still managing usage costs on the backend.
The key, he suggests, is predictability and transparency. While consumers hate surprise bills from utility companies, they're more accepting of usage-based models when they can track their usage and receive notifications about approaching limits.
Looking Forward
The speed of this transformation is remarkable. Woody notes that the shift toward usage-based pricing is happening much faster than the subscription movement that Zuora helped pioneer in 2010. As AI continues to permeate every aspect of software, we're likely to see even more innovation in pricing models.
For consumer software companies, the path forward might be a hybrid approach – combining the predictability of subscriptions with the flexibility of usage-based components. As ChatGPT and other AI-powered consumer services pave the way, they're making it easier for others to follow with similar models.
The future of SaaS pricing appears to be more nuanced and adaptable than ever before, with AI serving as both a catalyst for change and a model for new pricing paradigms. As Woody puts it, "Things are just changing at a rate of speed that is kind of hard to fathom."
🎯 Expert Insight
I've done maybe 10 pricing projects involving significant AI functionality this year. Here are some notes and how I’m currently thinking about it.
AI should be considered a 2-layer stack:
1. The AI compute 'fuel' (i.e. token pricing at OpenAI)
2. The AI solution (i.e. the value you add on top of AI)
The dilemma with AI pricing is that fuel is expensive relative to traditional compute, and solutions are immature. Plus, the key with any AI pricing model is that it needs to both work today AND tomorrow.
Today, companies are mostly focused on providing a low barrier to entry to get users onboarded in order to develop the solution layer and get data on behavior and cost patterns. This often means charging based on usage to ensure costs are covered, as usage patterns of customers are often unpredictable.
☝🏼 This is unsustainable as fuel costs will drop and 2025 customers will refuse to pay a price-per-token (or token equivalent) that is based on 2024 token costs. This is especially true for enterprise.
Tomorrow, companies will likely charge based on the outcome created by the solution layer and factor fuel costs into the use case. They’ll likely implement fair usage limits to protect against cost downside of over-usage.
Solving for Today → Tomorrow
Based on my experiences navigating these waters with clients, I've developed a transitional approach:
Accelerate adoption wherever possible. If you have a core non-AI product that's already monetizing effectively, consider your AI features as a strategic investment rather than an immediate revenue driver.
Practice transparency with customers. Explain that your pricing reflects both fuel costs and solution value, but be honest about the current weighting (typically 90% fuel, 10% solution in early stages).
Gradually shift the balance. Proactively reduce fuel pricing components, even anticipating future cost reductions before they materialize. Simultaneously, develop and communicate your solution value metrics independently.
Eventually eliminate explicit fuel pricing entirely. Integrate these costs into your solution pricing model, using fair usage limits to manage outlier scenarios—exactly as traditional SaaS products do today.
I don't claim to have all the answers, and my thinking continues to evolve with each new project. The most valuable insights often come from practitioners navigating these challenges in real-time across different contexts and industries.
Thanks for tuning in and see you next week!
Have thoughts on this post? I’d love to hear them. Hit reply or drop a comment.
Ahh, shucks, Ahrefs didn't make it :(