AI Is Upending SaaS Pricing

**Scott Woody** (0:00)
It used to be, the rule of thumb was, you don't change your fundamental pricing model more than once every five years. Agent Force, like literally Salesforce, one of the biggest companies in the world, changes fundamental pricing structure three times in the past 12 months. Now, that's like insane to me. I don't know how they retrain their sales team, and maybe it's going well, maybe it's not. But the point is that some of the largest companies in the world are moving at historical speeds. And so if you're the one sitting there, startups on the one hand are kind of like cannibalizing or rewriting your market, and you're sitting on a pricing change that you're going to sit and you're going to do a nine-month case study and then roll it out after six to nine months, you're going to be like a year and a half behind. And by the time you even go live, your pricing model is going to be outmoded.

**Martin Casado** (0:43)
Unless you're building, buying or selling enterprise software, that is. I'm talking about software pricing and how AI is turbocharging a shift to usage-based billing for enterprise products. Now, usage-based pricing isn't a new concept. We've been doing it with infrastructure for nearly two decades, but it definitely hasn't been the norm for most SaaS products. However, as AI shifts the value proposition from providing access to doing work, seat-based pricing makes less sense for many products, and usage-based pricing makes much more sense. Here to explain the rationale for this evolution and the intricacies of doing it well, from billing infrastructure to sales strategy, our Metronome co-founder and CEO, Scott Woody and a16z general partner, Martin Casado. You'll hear all they have to say on the topic after these disclosures. As a reminder, please note that the content here is for informational purposes only, should not be taken as legal, business tax or investment advice, or be used to evaluate any investment or security, and is not directed at any investors or potential investors in any a16z fund. For more details, please see a16z.com/disclosures.

**Scott Woody** (1:54)
Really, the kind of genesis of what we're doing, the way we think about it is like monetization infrastructure. So kind of take all the process of how a company generates value and captures it and turn it into software infrastructure. That's like ultimately what Metronome is all about. And if you're trying to think about it in terms of technology, it's roughly like data dog plus a billing engine is like what we literally do. Metronome actually starts from my time at Dropbox. So I worked there from 2013 to 2019, and I was in charge of essentially the engineering part of monetization, what we basically called monetization growth. So our job was to kind of do the pricing and packaging, build the front end of the experiences, launch new SKUs and add-ons. And our goal at the end of the day was to drive revenue growth of the business. So that experience really taught me the power of essentially monetization, pricing and packaging and helping grow a business. But there were kind of three challenges that my team ran into. So, the first was we would go do like a pricing experiment, let's go try to charge this set of users this different price. Instead of charging them $9.99, let's call them $11.99. Those pricing experiments would basically require someone to go write code in the billing system and frequently because that code was so fragile, it would take anywhere from a quarter to two quarters to get that pricing change live in production, which was insane because we could do the front end work in like a day, and so the cycle time of the experiment was like months. The second challenge we ran into was that we would do that change and then customers would get really confused. And the kind of root problem there was our billing system was essentially a script that ran once a quarter, once a month, once a quarter depending on the billing cycle. But customers were in the product every single day. They're looking at the price. They're looking at how much they're going to pay. And because, again, the bills were only computed like once every full moon, that meant that customers only got information out of the billing system like once a month. And so they'd find out about pricing changes after they'd already paid for the new price. So it's like this really challenging customer experience problem. And that really taught me that billing was really, honestly, more of a product surface than a once a month workflow. And then the third problem we ran into at Trollbox was, as we were customers of the billing system, but the data that came out of the billing system, basically, we didn't have access to. It kind of, you know, we'd write invoices, we'd get paid or not get paid, you know, they do these, like, collections processes, and then eventually it would flow into a general ledger, and then, like, some business analysts would do some work, and they'd tell us, oh, by the way, you made the company $6 million last quarter, something like that. That lag time of, like, experimental learning, it's like if you've ever worked in growth, it's like anathema. Basically, what it was saying is that our learning loop was roughly two quarters. That's way too slow for a business, and so I wanted to build a system that was fundamentally meant to solve the monetization problem.