Griffin Parry, CEO of m3ter – Interview Series

Griffin Parry

is the CEO and co-founder of m3ter. This is his second startup having previously co-founded and led GameSparks, a cloud services company acquired by Amazon in 2017, after which he spent 3 years working in senior product and field roles at AWS. He started his career in the media sector (Sky, News International) focused on digital strategy and digital product development, including launching and leading Sky’s online TV portfolio.

m3ter

is a SaaS platform designed to help companies implement and manage complex usage-based pricing by acting as a metering and billing infrastructure layer that sits alongside existing systems like CRMs and ERPs. It ingests raw product usage data, applies flexible pricing logic, and automates the entire quote-to-cash process, enabling businesses to generate accurate, real-time bills while reducing revenue leakage and operational overhead. By decoupling billing from core systems, m3ter allows companies to experiment with pricing models, launch new products faster, and gain deeper visibility into customer usage and revenue streams, making it especially valuable for modern software companies shifting toward consumption-based business models.

You founded and scaled GameSparks through to acquisition, then chose to start m3ter focused specifically on billing infrastructure and modern monetization. What drew you to this particular problem space for your second company, and how did your earlier founder experience influence that decision?

We’re a classic case of founders solving a problem we’d experienced firsthand.  At GameSparks, we had a modern monetization strategy — usage-based pricing — because that worked for the kind of business we were in (cloud infrastructure).  It was key to our success, but it also caused a lot of operational and GTM pain.  Then at AWS, also a cloud infrastructure business, albeit a much bigger one, we saw they had the same problems.  We also saw how much effort they put into solving them because it was critical for their business.  We came to realize that in a usage-based world, billing infrastructure is a strategic capability that most companies couldn’t develop, so we founded m3ter to change that.

AI-native products can have unpredictable infrastructure costs tied to inference, token usage, or model retraining. How should founders think about aligning pricing with value while protecting gross margins?

Traditional SaaS products generally had near-zero marginal costs of usage.  In other words,  the amount the customer used your product didn’t impact your cost to serve.  That’s not true of AI products because their usage drives costs such as token consumption.  If your prices are fixed, that means your gross margins can vary significantly per customer depending on their usage.  That, in turn, makes usage-based pricing strategies almost inevitable:  it aligns revenues with costs and stabilizes gross margins.

As AI becomes embedded into existing software categories, do you expect most companies to layer usage components onto subscriptions, or do you see entirely new monetization frameworks emerging?

I don’t expect anything entirely new – just a reinvention of pricing models we’ve seen before. You’ll see the full spectrum, from pure subscriptions to outcome-based models. But the biggest cluster will be hybrid: fixed recurring elements for predictability, combined with a variable metric that works for both customers (they associate it with success) and vendors (it is aligned enough with costs to protect their margins).

There’s growing discussion around outcome-based pricing in the AI era. Where do you see real traction emerging, and where do you believe the model becomes too complex to implement effectively?

The challenge with outcome-based pricing is attribution – for it to work, a measurable outcome needs to be unambiguously driven by the vendor’s product.  Sometimes that’s possible – payments is an example, where providers take a share of the transaction, and that seems fair.  But in my experience, these situations are relatively rare, and companies tend to fall back on pricing metrics that are more like proxies for value – for example, for an AI customer support agent, calls resolved without human intervention.  Again, there’ll be lots of solutions along the spectrum from usage-based, through value proxies, to outcome-based pricing – it depends on the use case.  What they all share in common is that something needs to be counted and have pricing applied to it, which is where m3ter comes in.

When defining value in AI-powered products, what practical metrics should companies focus on as realistic proxies for outcomes?

This is a difficult one to answer, because it is very use-case specific.  There are some “always” considerations – is the metric simple, predictable, associated with value, and well enough aligned with costs to serve?  But the metric itself depends on what the product does.  “Tokens used” works for an LLM model.  “Documents processed” works for contract analysis.  “Queries executed” works for enterprise search.  “Conversations handled (without human intervention) works for customer support.

What are the most common operational and technical challenges companies face when shifting from subscription-only models to hybrid or usage-based pricing?

The key pain points are around revenue leakage, poor customer experiences, and a lack of pricing agility that hamstrings Product and Sales.  The causes are rooted in the wrong operational foundations.  The key (new) capabilities needed when shifting from subscription-only to hybrid or usage pricing are usage data processing, advanced (and continuous) bill calculation, and automated connections between CRM, billing, and ERP systems.

Many enterprises are deeply committed to systems like Salesforce and NetSuite. How does m3ter modernize monetization infrastructure without forcing companies to overhaul their existing stack?

Established quote-to-cash tooling like Salesforce and NetSuite assumes a world of subscriptions.  That doesn’t mean they can’t work well for modern monetization approaches – you just need to fill critical gaps, which is what m3ter does.  We focus precisely on what’s missing: usage-data processing, advanced rating, and the automation of data flows between quote-to-cash systems.

Revenue leakage is often underestimated. How significant is this problem in modern SaaS businesses, and what typically causes it?

Revenue leakage is value that’s been earned (you’ve sold it and delivered it) but which hasn’t been collected because of billing inaccuracies – your bills don’t fully capture all customer usage, or don’t apply the right commercial terms.  It’s a big deal – PwC’s Revenue Integrity team estimates it at 4-7%, and the more complex the pricing, the more likely it is.  The root cause comes down to systems and controls: not capturing usage data effectively; not having automated connections between sources of truth for pricing and the bill calculation mechanism; and the bill calculation mechanism not being sophisticated enough to handle complexity (for example, relying on spreadsheets).

How does greater pricing flexibility influence product innovation and sales strategy within software organizations?

Simple – the more pricing agility you have, the faster you can ship new products, and the more easily you can adapt pricing to the needs and desires of your customers, including in private pricing deals that help Sales win.  It’s a strategic capability for the business.  But you can’t have flexibility without automation and control.  Otherwise, you get billing errors, revenue leakage, and compliance challenges.

Looking ahead, do you see AI playing a role in dynamically optimizing pricing models in real time, and what would need to be in place for that to work reliably at scale?

I’m certainly very excited about the potential for AI in price optimization.  But I’m less persuaded about the real-time aspect, at least for software-as-a-service or solution-as-a-service businesses.  If you’re selling hotel rooms or airline seats, dynamic pricing works because it’s a one-off transaction.  But B2B software vendors want customer relationships that endure, and customers don’t want pricing to change unpredictably day-to-day.  So price optimization will focus instead on creating bespoke pricing for long-term deals – pricing designed to deliver the best outcomes for both the vendor and customer over multi-year relationships.

Thank you for the great interview, readers who wish to learn more should visit

m3ter

.