Clay’s pricing change sent shockwaves through the go-to-market tech world. The company voluntarily separated data credits from workflow credits, published its internal memo for anyone to read, and openly acknowledged the move would cost them revenue in the short term.

It was a bet that aligning price to value would reshape how their customers and, eventually, the entire industry think about SaaS pricing.

In this episode of the Get Paid podcast, host Manny Medina sits down with Karan Parekh, Head of Finance at Clay, and one of the architects behind that decision, to walk us through the full arc of how it happened.

A couple of years ago, Clay was primarily a data business. Customer enriched profiles across 150 aggregated vendors, exported the results, and moved on. Data credits made sense.

But the product evolved. The real value shifted to orchestration, qualifying inbound leads in real-time, routing prospects before they finished submitting a form, and automating research at enterprise scale. Customers started asking a reasonable question: why is data getting more expensive if what I’m actually paying for is the workflow on top of it?

Clay spent a year talking to over a hundred customers and agency partners, benchmarking against orchestration platforms, and testing pricing models before making the split. They cut data costs by more than half and introduced a separate credit for platform activity, the action that actually creates business outcomes.

The short-term bet was explicit: revenue would decline. The long-term thesis was that if customers could find five or ten things to do inside Clay instead of one, the math would overwhelmingly favor the new model.

The team debated collapsing everything into a single credit type. Customers found it confusing. If a credit represents data, why is it also paying for orchestration? Splitting into data credits and action credits created transparency. Data credits function like a wallet with generous rollover. Action credits function like a capacity ceiling, refreshing monthly.

A year ago, Clay was mostly PLG. Today, the business is approaching an even split between PLG and enterprise. Enterprise buyers want predictability, and Karan’s team delivers it through tight scoping: defining the use case, approximating credit consumption, and giving the buyer a concrete number.

The first use case lands, works better than expected, and the expansion motion becomes consultative: hackathons, on-site sessions, introductions to other customers.

Karan sees AI-era SaaS pricing converging toward cloud infrastructure economics. Storage is a low-margin commodity. You cover costs, but that’s not where you build a business. The value layer sits on top.

Today, most AI companies price tokens because they have to cover input costs. But tokens don’t represent value. Some workflows consume a few tokens and generate massive outcomes. Others burn through compute and produce nothing differentiated. The industry will eventually need two vectors: one for covering fixed costs and one for capturing value created.

Clay’s ambition isn’t to replace the CRM. Clay wants to be the system of action. Wherever your data lives, Clay pulls it in and helps you do something exceptional with it. Karan argues that Clay actually makes CRM data stickier by keeping it fresh and useful, rather than threatening the platforms that store it.

Karan pushes back on the narrative that AI will flatten go-to-market into pure automation. Before the printing press, the hard part was manufacturing books. Once manufacturing became trivial, the hard part shifted to having something worth saying.

AI will do the same to go-to-market: volume will explode, but the things that stand out will be driven by genuine creative insight with an increasingly short half-life.