Adaptive Pricing Models

AI Agents Don’t Buy Seats—Why Your Pricing Should Follow Suit In the past 12 months, a clear pattern has emerged: as AI systems replace manual effort with automated intelligence, pricing structures tied to “seats” no longer reflect the value customers receive. Pricing models have surfaced as a hot topic with every portfolio company at Mosaic Ventures and is top-of-mind for nearly every founder building applied-AI products. When one person and an AI agent can outperform an entire legacy team, charging per user starts to feel arbitrary; what matters is how much business impact the product delivers. Founders are experimenting with three broad approaches: 1. Usage-metered plans that bill against tokens, API calls, or minutes of inference time. These create a direct bridge between consumption and margin and nudge teams to track cost from day one. 2. Outcome-based pricing that charges per lead booked, ticket resolved, or document drafted—tying revenue to measurable results. It’s the software analogue of value-based care. 3. Hybrid “starter bundle plus runway” tiers: a predictable monthly fee with a healthy allowance of AI credits, then pay-as-you-go beyond that. This balances budget certainty for customers with upside capture for the vendor. Across our portfolio, a few design principles keep showing up: 1. Anchor on a metric the customer already tracks. If your product shortens sales cycles, price per opportunity accelerated—not per login. 2. Bundle enough volume to eliminate credit anxiety. No one wants to ration prompts. 3. Expose real-time usage. Transparent dashboards prevent bill shock and build trust. 4. Instrument cost early. Metering and billing belong in the product backlog, not the finance queue. 5. Plan for non-linear jumps. When a model upgrade multiplies compute, re-grade tiers before your gross margin does it for you. AI’s promise is to shift human effort from repetitive execution to higher-order creativity. If our pricing still counts bodies instead of business results, we undermine that promise. The companies that map price to outcomes—while keeping the buying experience refreshingly simple—will capture the most upside. I’d love to hear how others are managing the move from seats to usage and outcomes. What’s working, what still feels messy, and where do you see the biggest opportunities to innovate on pricing? #appliedAI #pricing #startups

Having founded two SaaS companies (Statista and ECDB), I've watched pricing models evolve from simple seat-based subscriptions to something far more nuanced. Today, I want to share my findings on what pricing model works for many SaaS companies, and what we’ve learned. Seat-based pricing means you buy x seats, use 70-80% actively, and everyone gets tool access. Simple. But the world has changed. Today, data flows through multiple channels, which means a seat does not reflect actual usage anymore. Data can now be accessed in various ways: 📈 Direct API integrations with BI tools 🤖 AI assistants answering ad-hoc questions 🖥️ Automated workflows pulling market data daily 🧑💻 MCPs (APIs for LLMs) enabling new use cases A single developer might automate queries for an entire organization. Ten analysts may share one dashboard but rarely log in. Why should they all pay the same? It doesn’t make sense. According to an OMR/hy study, usage-based pricing adoption in SaaS jumped from 31% to 67% in just two years. The reason? AI and automation are making per-seat models obsolete. When one employee can automate what previously required five, charging per seat doesn't reflect value delivered. The software’s true value comes from enhancing efficiency, output, or outcomes, not the headcount. That is why we at ECDB are moving to a hybrid model: platform access + consumption credits. 👇 Here's our approach: 1. Platform tiers remain - You still choose a plan based on team size and features needed. 2. Credits introduced - Each plan includes base credits for downloads and light API usage. Heavy automation requires add-on credit bundles with volume discounts. 3. Fair pricing across channels - Whether you access a data point via xls, API call, or AI query - same credit cost. No more arbitrary pricing based on how you access the data. We found that this model works best for us right now. I welcome feedback from our customers, other SaaS founders, and industry experts. Are you seeing similar shifts in your products?

Seat-based pricing is dying faster than most CEOs realize. 12 months ago: 50% of B2B SaaS used flat or seat-based models Today: Down to 27% Projected 2027: <10% The shift? 41% of companies are now using hybrid pricing models. At SaaS Metrics Palooza, I walked through why this is happening, and the framework we use to build hybrid models that actually work. Here's the reality: AI changed who does the work. It's not just humans anymore. It's systems completing tasks, generating summaries, resolving tickets, approving requests, all autonomously. So how do you charge for that? By the person? The product? The outcome? Answer: It depends on who's doing more of the work. The BAM Framework - Base, Allowance, Meter These 3 layers create 5 different hybrid pricing models: 1/ Access Tiers (Base only) - Different AI complexity per tier, fair usage policy 2/ Flat + Unit (Base + Meter) - Platform fee + outcome-based hybrid pricing 3/ Flat + Limit + Unit (All 3 layers) - Most popular hybrid, includes volume + overages 4/ Credit-Based (Base + Allowance) - PLG favorite hybrid, but don't fall into cost-plus trap 5/ Pay-As-You-Go (Meter only) - Pure consumption, AI infrastructure The pattern from 400+ transformations: Companies using Flat + Limit + Unit (Model 3) are seeing the most traction. Why? It gives customers predictability (base fee + included volume) while capturing expansion value (usage after allowance). But, and this is critical, you need safety mechanisms. → Predict usage (dashboards, estimators) → Prevent surprises (match reset timing to contracts) → Protect customers (caps, true-up options, don't penalize usage) The companies winning with hybrid pricing aren't just picking a trendy model. They're designing the model that fits how their AI actually creates value. Based on Tremont's research: Hybrid models are capturing 3-4x more expansion revenue than traditional seat-based pricing. The question isn't "should we go hybrid?" It's "which hybrid model fuels OUR growth?" Which model fits your product? Person doing the work or product doing the work?

I've seen countless companies relying on outdated models or gut instincts for price changes. That often leads to tactical, knee-jerk pricing, missed profits, or constant battles to justify pricing & promotional plans to supply chain partners. I just recorded a quick video explaining exactly how we combine four different approaches to model elasticity accurately: 1. Double Machine Learning (DML) - Delivers a robust causal estimate by predicting sales and price from confounders, then regressing the residuals. - We typically build one DML model per SKU. In our experience, this often reflects real-world behavior best. 2. Log-Log regression models - It is simple and interpretable - perfect if you have lots of historical data, a high volume of transactions, or price variation. - The log price coefficient directly translates to elasticity. It is quick to implement, though it often oversimplifies and is not a good method for B2B. 3. ElasticNet - A regularized linear model balancing Lasso and Ridge methods. - If you have many variables, such as our promos, competitor promos, distribution, comp distribution, etc., it helps prevent overfitting. 4. Random Forest - Handles non-linearities pretty well without having to do complex data engineering. - We use price perturbation, simulating different price points to see how predicted demand changes, thus estimating implied elasticities. In the video, I also share how we compare the four methods, track metrics like RMSE or MAPE, and deliver scenario-based recommendations about price, promotions, and competitive moves, helping you go from reactive to proactive pricing. The real payoff is that you can: 1. Proactively manage pricing: estimate the impact of competitor actions and optimize your strategy. 2. Maximize promotional ROI: estimate what truly drives incremental volume vs. what's wasted spend. 3. Earn insights-backed credibility: support your pricing with robust elasticity metrics that show retailers how you got to your recommendations. I'd love to hear your thoughts. If you're ready to take a deeper look at these elasticity models (complete with a whitepaper, sample code, and practical examples), check out the comment section for links and more details!

If you’re still pricing #GenAI like SaaS, you’re not “innovating” — you’re gambling with your margins. AI-enabled business models are just emerging, but a recent article from Bessemer Venture Partners, "AI Pricing & Monetization Playbook" (in the first comment) nails the core shift: AI doesn’t monetize access; it monetizes outcomes — in a world where every token (and human-in-the-loop) has a real COGS line item. Practically speaking, start with the business model you’re really building: Copilot vs. Agent vs. AI-enabled Service → different economics, different charge metrics. Then pick a charge metric as a strategic choice (consumption → workflow → outcome): tighter value alignment means you’re taking on more cost risk. Next, use hybrid pricing (base + usage/outcome tiers) to balance predictability with upside. Finally, test value-first, then “find the price through friction” (if it’s an instant yes, it’s probably too low). Most importantly, treat pricing as your operating model: it shapes sales motions, CS incentives, what you measure, and how you scale from 10 to 1,000 customers. This resonates strongly with what I’ve been seeing in my research and in the classroom at The Wharton School: in AI-enabled business models, pricing isn’t a “packaging” decision—it’s where strategy, unit economics, and organizational design meet. #AI #GenAI #Pricing #Monetization #BusinessModels #UnitEconomics #GoToMarket #SaaS #Wharton

$7,225 for one day of coding. And Cursor isn't even the worst example. Replit's margins went negative. Anthropic throttles its best users. I mapped pricing across 50 AI startups. Six distinct patterns emerged. The core tension: traditional SaaS has near-zero marginal cost per user. AI products pay for compute on every interaction. A casual Claude user costs pennies. A developer running Claude Code all day costs tens of thousands per month. Your best users are your most expensive users. That tension is breaking every pricing model in the market. Cursor charged a flat 500 requests/month. Worked fine until users leaned into multi-step agent workflows. They switched to credit pools. One developer burned 500 requests in a single day. The plan description changed from "Unlimited" to "Extended" twelve days after launch. Replit grew 15x in ten months ($16M to $252M ARR). But they were buying revenue with compute. When they launched a more autonomous agent, margins crashed to negative 14%. They had to invent "effort-based pricing" mid-flight. Anthropic played it differently. Their $17/$100/$200 tiers map to genuinely different user personas, not volume bands. A casual user and a Claude Code developer are different products with different willingness to pay. The lesson across all 50 companies: before you set any price, pull the cost distribution. What does your P10 user cost? P50? P90? If the ratio exceeds 10x, flat pricing will break. In AI products, it almost always exceeds 10x. Full guide with all 6 models, 4 case studies, and a decision tree: https://lnkd.in/gdKaQSMk

I have spent years in the highs and lows of the consumer goods industry but never seen a pricing climate quite like this. Manufacturers are getting squeezed from every direction-tariffs, skyrocketing raw material costs, and relentless supply chain disruptions. The old playbook of raising prices to cover costs? That’s dead. Why? Because consumers are feeling the pressure too. A 2024 Nielsen report makes it clear: today’s shoppers are scrutinizing every dollar they spend, and brands that aren’t strategic about pricing risk losing market share fast. Here’s what I’m seeing from top CPG brands that get it: 1️⃣ Walmart is investing heavily in AI-driven pricing models to keep costs competitive-e-commerce now makes up 18% of total revenue. 2️⃣ PepsiCo is doubling down on pack-size innovation, offering smaller, affordable options to maintain volume without excessive discounting. 3️⃣ Luxury brands are using price elasticity models, testing demand thresholds before rolling out increases-avoiding consumer pushback. 4️⃣ Supply chain resilience is non-negotiable. Companies are shifting manufacturing away from China, despite short-term cost spikes, to avoid future geopolitical risks. The smartest brands aren’t just reacting. They’re rethinking. They’re moving toward Revenue Growth Management (RGM) frameworks that help them: ✅ Optimize pricing and promotions (because blanket price hikes are a losing game) ✅ Focus on margin-smart growth, not just revenue ✅ Leverage data analytics to make smarter, faster pricing decisions Brands that don’t evolve risk eroding profitability or pricing themselves out of the market. CPG leaders who master strategic pricing, operational efficiency, and consumer-driven value creation will own the future of this industry. Are you adjusting your strategy, or just reacting to rising costs? Because in 2025, only the most adaptable brands will win. #CPG #FMCG #PricingStrategy #RevenueGrowth #ConsumerGoods

Saas pricing is evolving. Many companies try figure out their pricing by copying competitor pricing or debating between conventional options. This misses a couple of key points. Your pricing needs to reflect two critical factors: F͟r͟e͟q͟u͟e͟n͟c͟y͟ Daily active users aren't the same as monthly active users. Why would you price them the same? Accounting software used by your CFO every 2 weeks vs design software used by a junior designer every day meets a different need. B͟u͟s͟i͟n͟e͟s͟s͟ ͟I͟m͟p͟a͟c͟t͟ A bug in your CFO's accounting software is potentially catastrophic. A glitch in your designer's tool? More of an inconvenience. Your pricing should reflect this risk and impact difference, not your opinion of "value". With AI and no-code tools making it easier to build interfaces (I whipped up an API integration in 20 mins the other day), we're seeing a fundamental shift in pricing models. The data backs this up: • 46% of SaaS companies now use hybrid models (subscription + usage) • Only 15% are purely usage-based This isn't all roses though: • 66.5% of IT leaders report unexpected charges from AI/usage-based pricing • Companies underestimate their SaaS spend by 304% So there's certainly hiccups in the transition period. So evolution is moving from seat-based to: • API-based pricing (like OpenAI's token model) • Usage-based pricing (like Snowflake's compute credits) • Task completion pricing (like Intercom's resolution-based pricing) Get this wrong and two things happen: • Churn - customers move to solutions that better match their needs Your pricing needs to align with how customers extract value. • Lost revenue - you're leaving money on the table If your pricing doesn't reflect real value delivered, you're missing out. Because it better reflects the varying value different customers get from the platform. Make sure to run qual & quant research to understand your users and the market rather than speculate. The future isn't about seats - it's about value delivered. As AI capabilities expand, expect even more granular and output-driven pricing strategies.

Selling to ENT without changing your pricing model is like showing up to a black-tie event in flip flops. MM pricing models don’t survive in enterprise sales. Why? Because selling 1,000 licenses to an enterprise isn’t 20x harder than selling 50 - but if you don’t adjust your pricing strategy, it will be 20x more painful. Enterprise buyers don’t think in per user terms. They think in budgets, forecasts, and cost centers. They want predictability, not a CPQ nightmare where they’re adjusting seat counts every quarter. If you’re moving upmarket, here’s how to avoid looking like a tourist at the grown-ups’ table: 1. Kill per-user pricing for large accounts. Enterprise CFOs see per-user models as a ticking time bomb...every new hire adds cost. Instead, sell in committed tiers, annual volume contracts, or all-you-can-eat licenses. - Instead of “$50 per user, per month,” structure it as, “$X for up to 1,000 users.” - Price for usage, not headcount - think storage, API calls, transactions, etc. 2. Enterprise doesn’t “expand naturally.” Build in expansion from day one. For MM, you can land small and grow. Enterprise doesn’t work that way. - Ramp pricing: Year 1 at 60%, Year 2 at 80%, Year 3 at 100%. Predictable growth, no CFO freak-outs. - Auto-expansion clauses: If usage exceeds X%, licenses auto-scale. Protects you from procurement pulling a “we’ll just add seats later” stunt. 3. Enterprise buyers expect to “win.” Give them a win - without losing. These buyers are trained to negotiate. They want a lower per-unit cost, but they’ll commit bigger dollars to get it. - Introduce an ENT Rate...lower per-unit cost, but higher minimum commit. CFOs love “efficiency,” and you get more ARR locked in. - Structure custom packaging that makes them feel special. Limited access to beta features, priority support, or bundled services. Want to win in enterprise? Stop selling like an SMB rep. Price for scale, control the expansion, and let procurement “win” on terms that make your CFO smile.

Sharing this from years of working across Quant, Risk, and Model Development 📊 📊 One thing people outside the industry rarely realize is that different companies use completely different pricing models for the same product. There is no universal model. Each firm chooses a modeling framework that fits its business, liquidity needs, risk appetite, and computational limits. For example: High-frequency trading firms lean toward models that are extremely fast. They prefer closed-form formulas, simplified volatility surfaces, fast-converging lattices, or approximate PDE solvers. Accuracy matters, but speed and stability matter even more, because they reprice thousands of instruments per second. Investment banks rely heavily on models that are robust under stress, not just fast. Their goal is consistency across trading, risk, and regulatory reporting. Buy-side firms such as hedge funds or asset managers can afford more complexity. Their priority is modeling realism because a small improvement in accuracy can materially change PnL. Clearing houses and exchanges focus on transparency and reproducibility. They prefer lattice models, finite-difference PDEs, and well-controlled Monte Carlo frameworks that can be audited and validated easily. Everything must be explainable, stable, and fully documented. There is no single perfect model. There is only the model that fits the purpose, the horizon, the liquidity, and the risk profile of the firm. That’s the beauty of this field. Two companies can price the same option using entirely different engines, and both can be right; as long as the assumptions, calibrations, and risk frameworks are consistent.