B2B SaaS has begun a shift to delivering functionality through agents. This is likely bigger than the shift to the subscription model that started more than two decades ago. What will this mean for pricing? Let’s share ideas.
2 Likes
I think what’s interesting is that the price is compared to current human equivalents.
Lovable went from 0-$10M in ARR in 2 or 3 months.
Same with Replit $30M in ARR in a quarter or so.
That’s b/c they are replacing fullstack devs…
I went from paying $20 to $200 in 19 days on Lovalable.
Still considering it a bargain as equivalent outsourced dev would have been at least $4000 or more …
1 Like
I would love to hear @Arnon’s take here. He and Manny Medina are building something awesome in this space 
Totally agree that agents are a game-changer. A few things I’m keeping an eye on:
- Agents enabling outcome-based pricing
- Agents making humans a premium feature (e.g., Agent handles low-complexity work, human handles nuance)
- Agents focused within specialized verticals (very bullish on this)
Manny and I are talking to companies about getting their agents paid somewhat like humans in compensation structure:
You can pay a “flat fee” almost like a retainer, but you want to give them a bonus for completing tasks well.
Think about things like “I just booked you a meeting with an ultra high value client” - pay that bot an extra $20 for succeeding, as opposed to just pre-purchasing a set of 200 leads per month.
1 Like
OpenAI made an important move when it tested the idea of an agent paid $20,000 per month. That reframed the anchor price for agents. B2B software is generally expected to generate anywhere from 10X to 20X the value of the cost, that is $200,000 to $400,000 for a $20,000 agent. A high hurdle for many of the agents coming to market. On the other hand, a knowledge worked is often expected to bill out at 3X to 5X of her fully loaded cost. That would be $60,000 to $100,000 per month in value. Not trivial but a lower hurdle. Which measuring stick will these agents be held to?
1 Like
Great question @Steven_Forth! So curious to see where this lands. Hadn’t thought about it in those terms before.
Just off a fascinating call with the CEO of Brighthive (AI agents for data management). She has a compelling story to tell about how she has designed her pricing. We will include a case study in the report on agenticAI that we are working on with Maxio. The short version, tokens and credits just confuse the buyer and don’t help anyone understand value. Instead, Brighthive prices on the type of agent (the value of different types of work is different), the work being done by the agents, and what has been delivered. There is a dashboard to show all of this inside the agent management console.
1 Like