← All insights

Billing on relationships, not minutes

How you price a product tells your customer what you actually optimize for. Most of the voice AI industry prices per minute — and per-minute pricing quietly puts the vendor and the customer on opposite sides of the table.

The problem with per-minute

When you bill by the minute, your revenue goes up when conversations get longer. But a longer conversation is usually a worse outcome — it means the system was slow, confused, or made the user repeat themselves. Per-minute pricing rewards friction. It also races to the bottom: minutes are a commodity, every competitor undercuts on price, and the number that matters most to your customer — whether their users come back — isn't in the bill at all.

Pricing on Monthly Active Relationships

VoiceQuest bills on Monthly Active Relationships (MAR): the count of distinct, remembered relationships your product is actively maintaining in a month. It's a simple idea with a sharp consequence — our revenue grows only when your users keep coming back. If a relationship goes cold, we don't get paid for it. That puts us on the same side of the table as you: both of us are trying to maximize retention, not talk time.

Why it fits the memory layer specifically

MAR isn't a gimmick bolted onto a per-minute product — it's the natural unit for a memory layer. What we provide is precisely the ability to maintain a relationship over time. Billing on the number of active relationships prices the thing we actually deliver. For device makers where a "monthly active" meter is awkward, the same logic maps onto per-device licensing: you pay for the capability shipped, not for how much your users happen to talk.

What this means for you

Across all three, the principle is the same: we win when your users stay. That's the incentive you want from the layer that owns your customers' memory.

See the pricing →