⭐ At defy.vc we've been exploring the evolution of pricing in an AI-first world. 🚀 AI-first companies are shifting away from traditional SaaS models and embracing more dynamic, transactional pricing strategies to capture larger market opportunities—especially those tied to labor spend. 🔑 Why the Shift? With AI automating workflows and replacing manual labor, businesses are exploring new ways to price based on input or output rather than the typical subscription model. The goal? Align with customers' specific ROI and cost structures while tapping into industries where labor is a significant cost. 💡 Key Pricing Models for AI-First Products: 1. Fixed Cost (SaaS) – Ideal for daily-use cases, like AI generated marketing collateral or compliance checks. 2. Input-Based – Charges by consumption (e.g., data processed or queries made). Think of AI-powered discovery tools for law firms. 3. Output-Based – Pay for specific outcomes (e.g., automating administrative tasks in healthcare). 4. Hybrid Models – A mix of subscription and usage, giving flexibility without sacrificing predictability. 💬 When to Choose Which? It all comes down to how often your customers use the product, their existing budget for labor vs software, and how aligned your pricing is with the value they perceive. As the AI market grows, so does the opportunity to experiment with pricing that reflects real-world savings and efficiency improvements. The right model could unlock new revenue streams—and new value for your customers. Curious how to navigate this shift? Dive deeper into the pros, cons, and use cases of each pricing model. 🌟 Link to the full post in the comments. Thanks to Ximei Li for the collaboration and thought partnership! #AI #PricingStrategy #TechInnovation #SaaS #AIProducts #StartupStrategy #BusinessModel
1000% agree. I think transactional models, in vogue or not, will be the way to capture the highest total revenue moving forward. AI is already accelerating the commoditization of software so SaaS pricing will be a race to the bottom and not worth the benefits. To be clear, I think SaaS TAM will grow but ACVs will shrink. On transactional pricing for software- Ramp is a great example - why pay for expensify, bill.com etc when you get the software for "free" given subsidized by the transactional revenue? I'd also bet the Ramp ACV is 10x the same customer's ACV for expensify or bill. Win win.
I have liked the hybrid models that charge a recurring platform fee that guarantees a certain minimum usage e.g. up to X terrabytes of data and then charging for overages, which is similar to another defy company PuppyGraph. That way the customer has predictability in the bill and when usage expands, the billing has a way to grow with the client's growing business. Thanks for outlining the four most common pricing models you are seeing out there!
I remember an observation a few days ago that "serverless" isn't an engineering pattern, it's a pricing strategy. We made hosting serverless, it sounds like we're about to do the same to software.
What an enlightening post! It's fascinating to see how AI-first companies are reshaping the landscape with these innovative pricing models. The idea of aligning pricing with customer ROI and labor costs is a game changer. I'm curious about real-world examples where these models have been successfully implemented or any challenges encountered. Looking forward to learning more!
The shift from flat SaaS to usage or outcome-based models is a smart way to tie value directly to ROI. Hybrid models, in particular, seem like the sweet spot for balancing customer trust and scalability. Excited to see how this evolution reshapes market dynamics!
And to implement and refine all these pricing models, use Lago’s open source billing platform :-) cc Anh-Tho Chuong
Great article Medha. Definitely a big evolution for software / SaaS as transactional models become drivers of many AI business models. Thanks for sharing.
Interesting and relevant to Kathie Yang's work. Thanks for sharing!
General Partner @ Defy
2mohttps://makecents.substack.com/p/how-to-price-your-ai-first-product