Marcus Webb raised Northfield's Series A in mid-2024 with strong metrics: $2.1M ARR, 85% gross retention, seven enterprise logos. The pitch worked. The pricing model that powered those metrics was already becoming a liability, and Webb knew it before the wire landed.
Northfield sells workflow automation software to mid-market operations teams. The original pricing was $299 per seat per month — straightforward, easy to sell, easy to understand. The problem was that the value Northfield delivered was not correlated with seats. Customers with five power users and 200 passive users were paying the same as customers running the full platform across a department of 50. The active users were getting enormous value. The passive seat holders were justifying cost at renewal and not winning the argument internally.
The signal that something was wrong
"Our NRR was 108%," Webb says. "Which sounds fine until you look at what it was made of. We had a handful of customers expanding aggressively and a long tail of customers flat or churning. The aggregate number was hiding a distribution that was really telling me two things: we had some customers who loved the product and some customers who had bought it wrong."
Buying it wrong, in Northfield's case, meant buying seats for an entire team when only a fraction of the team was operating the platform actively. The per-seat model gave procurement a clean comparison (cost per person) without surfacing the more relevant metric (value per workflow automated). Customers who evaluated on a per-seat basis had a fundamentally different relationship with the product than customers who evaluated on workflow ROI, and they were making renewal decisions from the same frame.
The rebuild
Webb spent six weeks before making any external changes doing internal analysis: which customers were using the product at the depth that correlated with retention and expansion, and what was the common characteristic of those customers? The answer was workflow volume — the number of automated workflows running per month — rather than seats. Customers above 500 active workflows per month were retaining at 97% and expanding at an average of 40% annually. Customers below 200 workflows were churning at 28%.
"The pricing was selecting for the wrong customers," Webb says. "Per-seat pricing attracts buyers who think about the product as a collaboration tool. Usage-based pricing attracts buyers who think about it as an automation platform. Those are different people making different decisions for different reasons, and we needed the latter."
The new pricing moved to a base platform fee plus workflow volume tiers. Existing customers were grandfathered for 12 months with a clear migration path. New customers saw the new model immediately. Sales was retrained on a new discovery motion that surfaced workflow volume potential in the first conversation rather than seat count.
The result
Eight months after the pricing rebuild, Northfield's NRR had moved from 108% to 127%. Gross retention improved from 85% to 91%. Average contract value for new customers was up 34%. The customer mix shifted toward larger, more operationally complex buyers who were evaluating on workflow ROI rather than per-seat cost comparison.
"The hardest part was not the pricing change," Webb says. "It was convincing myself that the short-term hit to new customer closes — because usage-based pricing takes longer to evaluate than per-seat — was worth the long-term improvement in customer quality. It was. But that is a harder conversation to have with your board in month two than it is to have looking back from month eight."
Pricing is not the number. Pricing is the frame through which customers decide whether the product is worth keeping. We changed the frame and got different customers. That was the whole point.