Per seat pricing is dead – this is the future of SAAS business

The SaaS pricing model is changing as AI agents become central to business operations. Ulrik Lehrskov-Schmidt, founder of Willingness to Pay and author of "The AI Pricing Roadmap," explains that companies like HubSpot have already tripled prices by moving to credit-based usage, impacting existing customers.

This shift means that AI agents now act as gatekeepers, making initial vendor selection critical because they rarely switch. These Big Ideas are drawn from Ulrik Lehrskov-Schmidt's interview on SaaStr AI.

• HubSpot's pricing shift and growth concerns. HubSpot's new pricing model has drastically increased costs for existing customers, with one example showing a jump from $1,300 to $3,700 per month. This increase is largely due to a shift from seat-based to credit-based usage for workflow actions. Ulrik Lehrskov-Schmidt suggests that such aggressive price adjustments are often a strategy by companies facing declining growth to compensate for shortfalls in new customer acquisition. ▶ 16:35

• AI agents as vendor gatekeepers. AI agents are becoming key decision-makers in vendor selection. Once an agent 'picks' a vendor, it tends to stick with that choice, similar to human inertia but for different reasons. For example, an agent might choose a tool like Higgsfield and not switch unless a significant issue arises, making initial adoption by agents critical for long-term success. ▶ 5:11

• Speed as the new moat. The ability to rapidly adapt to new AI model changes, capabilities, and pricing structures is now the primary competitive advantage in SaaS. This 'speed as a moat' contrasts with traditional moats like product features or established pricing, as companies that can iterate and integrate faster will outpace competitors. ▶ 29:15

• Balancing AI agent autonomy. Determining the right level of autonomy for AI agents is a complex challenge. While too little autonomy can be exhausting for human oversight, excessive autonomy poses significant risks, such as an agent autonomously ordering unapproved items or emailing an entire customer database without proper authorization. This balance requires careful consideration and the implementation of guardrails.

• Security risks of custom CRMs. Building custom CRMs or applications for sensitive data, a practice sometimes referred to as 'vibe coding,' introduces substantial security vulnerabilities. Unlike established platforms such as Salesforce or Shopify, custom builds often lack the robust security hardening necessary to protect personally identifiable information (PII). An example cited is the WorkOS incident, where their custom CRM was breached by an open-source penetration testing tool. ▶ 59:15

• The rise of 'vibe building'. The trend of 'vibe coding' internal applications is largely over due to security risks and the impracticality of maintaining custom systems. However, a new wave of 'vibe building' is emerging, focusing on creating custom agents, headless systems, and user interfaces. This shift indicates a future where organizations will extensively build their own agents to address specific enterprise needs. ▶ 1:04:53

• AI agents in sales processes. AI agents can play a crucial role in sales by automating tasks that typically consume significant human seller time. For instance, an agent like '10K' can co-write deals, generate custom decks with competitive analysis, price custom requests, and manage procurement details. This allows human sales teams to focus on strategic interactions and relationship building.

• Disruptive potential of agent collaboration. The capacity for AI agents to communicate and collaborate, or 'swarm,' is a highly disruptive development. This allows agents to call upon each other in real-time to brainstorm or solve problems, as demonstrated by Muse and 10K identifying a user's time constraint. However, this increased autonomy necessitates careful management to prevent unintended actions or security risks. ▶ 47:55