At a recent team meeting a discussion broke out about pricing in generative AI driven businesses. It turned out that as a team we’d been getting a lot of questions from our portfolio companies around pricing — more often than in the past.  We were a bit surprised because while pricing is integral to value capture/revenue in the long-run, it hasn’t historically been a key priority early on. This week’s conversation opened up my eyes to the potential power of new pricing models and aligned costs for customers in an age where a start-ups value prop may have shifted to doing the work vs. enabling the work.

📍 At Pre-Seed: Pricing Matters Less Than You Think

At the very beginning, pricing specifics are less important than proving you’re solving a real pain point and getting customers to fall in love with your product. Whether you’re building GenAI-powered software or a tech-enabled service, the goal is to land early design partners and get them to pay something—to have skin in the game.

Early revenue matters, but I don’t think the exact structure does. In fact, part of what venture capital does at this stage is subsidize your earliest customer base while you find true product-market fit.

Plus, constant iteration is normal—your ideal customer profile (ICP) and product itself may shift. Locking in rigid pricing too early can backfire. The one thing to consider: Are you “selling software” (where UX and interfaces matter most) —or “selling outcomes” (where service reliability and ROI are king)? This choice can (and should) influence how you plan to capture value.  In either case, one way of identifying willingness to pay in the longer term is by targeting existing budgets vs. finding new dollars or innovation budgets. (And the budget doesn’t have to be software spend, just dollars that are being allocated to a function today.)

📍 When Pricing Starts to Matter: Post-Seed

Once you raise your seed round and your first set of customers, pricing moves up the priority list.

You’ll need a clear strategy to:

• Build to critical mass of ARR for your Series A
• Show investors you’ve validated willingness to pay
• Prove there’s an economic GTM motion behind the product that aligns with potential annual revenue opportunity per customer.

📍 Real pricing power often kicks in post-Series A.

• Renewing customers at higher prices (of at least ‘full’ price vs. the early customer discounts) is easier than winning new ones.
• Upselling via new modules or agents, feature unlocks, or usage expansion becomes key.

If upsells aren’t structurally built into your product and pricing model, it gets harder to unlock multiple years of strong growth. Plan for it early.

🔥 Pricing Model Considerations

As you start to test pricing pre and post seed, here are few things to take into account:

1. Value to You (the Startup)

• Does potential customer budget over time justify the cost to acquire them?
• It may not have to pencil out in year 1, but if a customer will never pay more than $10K per year for that new customer onboarding agent, then it’s hard to scale a sales-led motion.
• If $10K starts as a wedge with lots of expansion upside, that might be fine.

2. Value to the Customer (the Buyer)

In vertical AI, pricing should reflect:

• Workflow transformation (time saved, errors reduced)
• Outcome improvement (revenue generated, risk reduced)

It makes sense to align pricing with outcomes (ROI), instead of model sophistication or data-processing costs. BUT if you anchor around business outcomes that you don’t fully control (for example, marketing leads or healthcare outcomes) then you can lose control of pricing.

In many industries, buyers are more comfortable with contractually defined pricing—regardless of how it’s calculated—because it’s easier to plan for and get approved. So even if you’re selling an LLM-powered product, it may make sense to structure pricing in familiar formats early on.

In more regulated industries like healthcare, legal, or finance, there’s also a strong case for platform-based pricing tied to integrations, compliance requirements, and long-term support. The key is to anchor pricing to clear, measurable value—without overcomplicating or tying it too tightly to outcomes outside your control.

🚀 Final Thought

Pricing strategy in vertical AI isn’t one moment, it’s a continuous discovery process. If you’re exploring pricing options and want to chat, we are here to help.