Keep pulling the thread on Mark Roberge.
The two primary competitive threats for a modern AI startup are the incumbent company and the underlying foundational model from providers like Anthropic or OpenAI.
Mark Roberge argues that product-market fit should be quantified by customer retention, not by revenue, customer count, or inbound demand.
Brett Taylor's new company grew from $0 to $150 million in ARR in six quarters.
Juben's startup, Roadrunner, has raised a Series A round led by Founders Fund.
A group of 35 tech CIOs from companies the size of HubSpot identified Configure, Price, Quote (CPQ) as the most broken problem in their organizations.
Juben's CPQ startup, Roadrunner, uses models like GPT-3.5 to allow account executives to define deal terms in natural language, which then reasons over price books and rule books.
A potential future enterprise architecture involves moving all company data into a data lake like Databricks and using a foundational model as a horizontal operating system to build custom agents on top.
Juben predicts that within one to two quarters, the market opportunity in CPQ will become obvious to everyone, leading to intense competition for his company Roadrunner from players like Glean.
The most common mistake Mark Roberge observed in startups was scaling revenue at the wrong time and pace.
Slack's leading indicator of retention was reportedly 80% of customers sending 2,000 team messages every month.
HubSpot's leading indicator of retention was 80% of customers using five or more features in the platform every month.
In Mark Roberge's experience, customer retention issues are more often caused by sales problems than by product deficiencies.