New Value Models, New Playbooks: Tactics AI-First Companies Use To Scale
By Medha Agarwal and Noah Lin
AI is redefining software’s role, enabling products that perform end-to-end work. This shift brings new pricing models, value metrics and GTM tactics.
At Defy, we’ve observed these four playbook stages that leading AI-first companies are using to scale.
Pricing and ROI: Selling value in a usage-based world
New pricing logic: AI products often use usage-based or hybrid pricing. That’s powerful, but less familiar. To succeed, teams must align pricing with outcomes and clearly articulate ROI.
Budget alignment: Unlike SaaS licenses or headcount, usage models need justification. For instance, Synthpop 1 charges per healthcare task automated — directly mapping cost to labor savings. This model resonates in labor-constrained industries.
Hybrid models for predictability: Blending tiered plans with usage minimums gives customers cost control while scaling affordably. For example: 10,000 credits at $500/month vs. 50,000 for $1,500. Lower unit costs reward growth.
Selling urgency: When pain isn’t acute, sellers should frame the cost of inaction. Ask:
- What’s the cost of staying manual?
- What happens if volume spikes?
- Can hiring solve this sustainably?
These questions qualify fit while creating urgency.
Discovery and qualification: Finding the right buyer
AI products require upfront investment, so qualifying buyers early is crucial.
Learn before pitching: Use discovery calls to understand how buyers currently tackle the problem. You can ask:
- What’s the current workflow?
- Who’s involved?
- Have you tried outsourcing or automation?
Position as labor alternative: Frame your product as a cost-effective way to avoid hiring. Ask:
- Is headcount or tooling the main constraint?
- Could this offset planned hiring?
Uncover real fit: Ask about competitors and hesitations around variable pricing. AI-first tools require real commitment — poor fit means wasted proof of concept and long sales cycles. Prioritize pain, urgency and organizational alignment.
Consultative selling: Guiding buyers through change
Once qualified, move from pitching to partnering.
Coach, don’t sell: Buyers often know the problem but lack a vision for solving it. Help them reimagine workflows and quantify the upside (speed, quality, reduced risk). Explain how your AI improves decisions — not just efficiency.
Build trust, not hype: Position your team as expert advisers. Highlight how competitors are adopting AI and frame your product as essential — not experimental. Focus on real problems, not futuristic features.
Co-create value: Buyers don’t want complexity. Understand their pain, then tailor a solution around it. When buyers feel heard and guided, they’re more willing to rethink their approach.
Proof through POCs: demonstrating real impact
A proof of concept isn’t just a technical validation — it’s also the key to proving value and earning trust.
Modern POCs = measurable outcomes: AI products tackle complex, variable tasks. POCs should reflect that — demonstrate consistent results across real scenarios, not just toy demos.
Structure for success: Successful teams scope tightly, set metrics early, and stay hands-on. Example:
- Origami Agents compares POC costs to hiring SDRs.
- Another AI platform focuses more on user enthusiasm and internal adoption than strict ROI, embedding in the workflow early.
Plan for conversion: Don’t wait until the POC ends to talk about the next steps. Begin commercial conversations midway, adjust pricing if needed, and ensure all stakeholders are aligned for expansion.
Final Word: Set the stage for long-term growth
AI adoption still feels experimental to many buyers. That’s why what happens after the sale matters just as much. Effective onboarding, early wins and long-term support are the foundation for retention and growth.
In our next piece, we’ll explore how AI-first companies succeed post-sale: from implementation playbooks to navigating internal resistance.
Medha Agarwal is general partner at Defy, where she partners with founders from the earliest stages. She previously spent seven years as a partner at Redpoint Ventures, backing early-stage companies including Whatnot, Tend, Proper Finance, LiveKit and Anvyl. Agarwal started her career at Bain & Co., founded two startups (Skedge.me and Roomidex), and invested at Bessemer Venture Partners. She studied social studies at Harvard, where she rowed varsity crew, and earned her MBA from Harvard Business School.
Noah Lin is an analyst at Defy, where he supports the full investment process, with a focus on sourcing and connecting with exceptional early-stage founders. He studied computer science and economics at Duke University. Before Defy, he was the first go-to-market hire at a high-growth Y Combinator startup, where he helped scale from early traction to repeatable enterprise sales.
Illustration: Dom Guzman
Synthpop is a Deny portfolio company.↩
Read More
Why More Startups Are Buying Other Startups In 2025
Fueled By AI, Fintech Ramp Raises $500M at a $22.5B Valuation Just Weeks After Last Raise
New Value Models, New Playbooks: Tactics AI-First Companies Use To Scale
By Medha Agarwal and Noah Lin
AI is redefining software’s role, enabling products that perform end-to-end work. This shift brings new pricing models, value metrics and GTM tactics.
At Defy, we’ve observed these four playbook stages that leading AI-first companies are using to scale.
Pricing and ROI: Selling value in a usage-based world
New pricing logic: AI products often use usage-based or hybrid pricing. That’s powerful, but less familiar. To succeed, teams must align pricing with outcomes and clearly articulate ROI.
Budget alignment: Unlike SaaS licenses or headcount, usage models need justification. For instance, Synthpop 1 charges per healthcare task automated — directly mapping cost to labor savings. This model resonates in labor-constrained industries.
Hybrid models for predictability: Blending tiered plans with usage minimums gives customers cost control while scaling affordably. For example: 10,000 credits at $500/month vs. 50,000 for $1,500. Lower unit costs reward growth.
Selling urgency: When pain isn’t acute, sellers should frame the cost of inaction. Ask:
- What’s the cost of staying manual?
- What happens if volume spikes?
- Can hiring solve this sustainably?
These questions qualify fit while creating urgency.
Discovery and qualification: Finding the right buyer
AI products require upfront investment, so qualifying buyers early is crucial.
Learn before pitching: Use discovery calls to understand how buyers currently tackle the problem. You can ask:
- What’s the current workflow?
- Who’s involved?
- Have you tried outsourcing or automation?
Position as labor alternative: Frame your product as a cost-effective way to avoid hiring. Ask:
- Is headcount or tooling the main constraint?
- Could this offset planned hiring?
Uncover real fit: Ask about competitors and hesitations around variable pricing. AI-first tools require real commitment — poor fit means wasted proof of concept and long sales cycles. Prioritize pain, urgency and organizational alignment.
Consultative selling: Guiding buyers through change
Once qualified, move from pitching to partnering.
Coach, don’t sell: Buyers often know the problem but lack a vision for solving it. Help them reimagine workflows and quantify the upside (speed, quality, reduced risk). Explain how your AI improves decisions — not just efficiency.
Build trust, not hype: Position your team as expert advisers. Highlight how competitors are adopting AI and frame your product as essential — not experimental. Focus on real problems, not futuristic features.
Co-create value: Buyers don’t want complexity. Understand their pain, then tailor a solution around it. When buyers feel heard and guided, they’re more willing to rethink their approach.
Proof through POCs: demonstrating real impact
A proof of concept isn’t just a technical validation — it’s also the key to proving value and earning trust.
Modern POCs = measurable outcomes: AI products tackle complex, variable tasks. POCs should reflect that — demonstrate consistent results across real scenarios, not just toy demos.
Structure for success: Successful teams scope tightly, set metrics early, and stay hands-on. Example:
- Origami Agents compares POC costs to hiring SDRs.
- Another AI platform focuses more on user enthusiasm and internal adoption than strict ROI, embedding in the workflow early.
Plan for conversion: Don’t wait until the POC ends to talk about the next steps. Begin commercial conversations midway, adjust pricing if needed, and ensure all stakeholders are aligned for expansion.
Final Word: Set the stage for long-term growth
AI adoption still feels experimental to many buyers. That’s why what happens after the sale matters just as much. Effective onboarding, early wins and long-term support are the foundation for retention and growth.
In our next piece, we’ll explore how AI-first companies succeed post-sale: from implementation playbooks to navigating internal resistance.
Medha Agarwal is general partner at Defy, where she partners with founders from the earliest stages. She previously spent seven years as a partner at Redpoint Ventures, backing early-stage companies including Whatnot, Tend, Proper Finance, LiveKit and Anvyl. Agarwal started her career at Bain & Co., founded two startups (Skedge.me and Roomidex), and invested at Bessemer Venture Partners. She studied social studies at Harvard, where she rowed varsity crew, and earned her MBA from Harvard Business School.
Noah Lin is an analyst at Defy, where he supports the full investment process, with a focus on sourcing and connecting with exceptional early-stage founders. He studied computer science and economics at Duke University. Before Defy, he was the first go-to-market hire at a high-growth Y Combinator startup, where he helped scale from early traction to repeatable enterprise sales.
Illustration: Dom Guzman
Synthpop is a Deny portfolio company.↩
Read More
Why More Startups Are Buying Other Startups In 2025