SaaStr Founder Shares Key Decisions For B2B AI Founders
SaaStr founder Jason shares key decisions for B2B AI founders, including re-benchmarking production workflows and reporting net new ARR in dollars. These decisions can help founders navigate pricing pressure and expanding TAM.
Original reporting · saastr.com
Original: “Jason’s Takes on This Week’s 20VC: A 72% Cost Drop on Our Own App, the $100M Acquisition Bar, and DoorDash’s Third Competitor”
Read “Jason’s Takes on This Week’s 20VC: A 72% Cost Drop on Our Own App, the $100M Acquisition…” at SaaStr →How this was written. Awesome B2B wrote this article with a language model from SaaStr’s reporting (model: @cf/meta/llama-3.3-70b-instruct-fp8-fast); it describes that reporting rather than adding to it. The original, linked above, is the full story.
Key points from SaaStr’s report
- B2B AI founders should re-benchmark their production workflows regularly to take advantage of falling model costs.
- Reporting net new ARR in dollars alongside the growth rate can provide a more accurate picture of a company's growth.
- Pricing pressure is a key consideration for B2B AI founders, and companies should plan for prices to fall 30-40% within a year.
- Agent-driven workflows are becoming increasingly important, and B2B apps should check whether an agent can use their product without a human logging in.
- The cost of serving persistent agents can be high, and companies should measure cost-to-serve separately for top users before setting a price.
SaaStr founder Jason recently shared his take on the latest 20VC podcast, discussing key decisions for B2B AI founders. According to SaaStr, these decisions include how to price for falling model costs and how to report growth when percentages are shrinking. Jason's learnings are based on his own experience with SaaStr Connect, the company's AI matching and recruiting app.
One key decision Jason highlights is the need to re-benchmark production workflows regularly. SaaStr Connect's costs fell 72% when Jason changed models and moved to a cheaper model mix. This change also resulted in faster and better output. Jason notes that founders who set their model mix once at launch and haven't touched it since are likely overpaying and may be underpriced by competitors.
Another important decision is how to report growth. Jason suggests reporting net new ARR in dollars alongside the growth rate. This is because percentage growth can be misleading, and net new dollars tell a more accurate story. For example, Anthropic's GAAP revenue grew from $5B to $11B to $17B over three quarters, but the percentage growth rate fell from 100% to 50%.
Pricing pressure is also a key consideration for B2B AI founders. Jason notes that a 30% discount from a competitor can halve a company's growth rate. He suggests planning for prices to fall 30-40% within a year and shifting pricing towards outcomes and workflows rather than model capability.
Jason also discusses the importance of agent-driven workflows. He turned off Canva after 12 years because SaaStr's agents didn't log into it. Instead, they use tools like Higgsfield and Opus that can be driven programmatically. Jason suggests that B2B apps should check whether an agent can use their product without a human logging in.
The cost of serving persistent agents is another key consideration. Jason notes that these agents can cost more than $20 per user per month, and power users can cost far more. He suggests measuring cost-to-serve separately for top users before setting a price.
Finally, Jason discusses the impact of agents on marketplaces and transaction fees. He notes that DoorDash's real competitors are Uber Eats and restaurants themselves, which can offer pickup or delivery options that avoid fees. Jason suggests that marketplaces should find the direct alternative an agent would route a customer to and price their value accordingly.
Questions and answers
How often should B2B AI founders re-benchmark their production workflows?
Founders should re-benchmark their production workflows every time a new small model ships, using their real tasks, not public evaluations.
How can B2B AI founders report growth in a way that accurately reflects their company's performance?
Founders should report net new ARR in dollars alongside the growth rate to provide a more accurate picture of their company's growth.
What is the impact of agents on marketplaces and transaction fees?
Agents can route customers to direct alternatives, avoiding fees and competing with marketplaces that take a commission.

