AI-Driven Business Models

Just published a framework for building an AI services company, going after the $20 trillion dollar industry powered by human-driven, low-tech businesses where legacy incumbents have built their brands on historical reputation and prestige rather than measurable performance. Here is the full framework & analysis: https://lnkd.in/gAEZ_Gq3 While many focus on AI making existing software more efficient, the true revolution is happening as AI pushes directly into domains previously exclusive to human experts: strategic negotiations, creative problem-solving, and high-stakes decisions. These critical domains have remained stubbornly resistant to software automation until now. We're witnessing a fork that will split the professional services landscape into two distinct futures: 🌑 Legacy Professional Services: - Dominated by established incumbents with century-old processes - Knowledge siloed by human experts with limited documentation - Services bottlenecked by human cognitive and time constraints - Premium pricing models based on artificial scarcity 🌕 Elite AI-Driven Professional Services: - AI-native firms delivering demonstrably superior outcomes - Expertise amplification across entire organizations - Services that scale beyond traditional human constraints - Value-based pricing tied to measurable outcomes This transformation represents perhaps the largest opportunity in the AI landscape today. For entrepreneurs and investors, the $20 trillion market of pure human expertise is now accessible in ways previously unimaginable. Let me know if you're building in this space. The future belongs to systems that truly learn from experience.

The AI business model is undergoing a fundamental transformation. For the last few years, the playbook was simple: put an AI wrapper on a SaaS product and sell it by the seat. That era is ending. The new wave of AI companies are moving beyond simple subscriptions and embracing a more sophisticated approach tied directly to value creation. Here’s what’s changing: 💰 From Seats to Spend: The most forward-thinking companies are shifting to usage-based and outcome-driven pricing. Think less about how many people use the AI and more about what the AI does. This includes new revenue streams like "agentic checkout," where AI agents complete purchases and transactions directly within a chat interface. The closer the AI is to the dollar, the more value it captures. 🎙️ From Text to Voice & Video: The interface for AI is becoming more human. Voice is mainstream (Sierra for support, Listen Labs for market research). The next frontier is video, where AI will see, understand, and interact with the world in real-time. The keyboard is no longer the only way to talk to a machine. 🤖 From Advisors to Actors: Early AI copilots gave advice. The next generation takes action. These agents aren't just suggesting what to do; they are executing complex workflows that directly impact the metrics that matter: boosting conversion, reducing average handle time (AHT), improving NPS, and cutting churn. This is about moving from passive assistance to active problem-solving. The common thread? A relentless focus on tangible ROI. We’re incredibly bullish on founders who understand this shift and are building companies that align their success with the success of their customers. The future of AI isn't just about intelligence; it's about impact.

This paper offers a comprehensive analysis of AI-driven business model innovation (BMI), identifying six key research dimensions crucial for understanding and advancing the field. 1️⃣ Triggers: Various factors trigger AI-driven BMI, including customer demand for AI-based solutions, technological advancements, data democratization, ecosystem developments, competitive pressures, regulatory compliance, and societal trends. These triggers drive companies to adopt AI to create new value propositions and enhance business model efficiency. 2️⃣ Restraints: Several barriers hinder AI implementation in business models. These include ethical concerns (such as algorithmic bias and misuse of AI), safety and security issues, legal and regulatory challenges, employee resistance, and the opaque nature of AI (the "black box" problem). These restraints can lead to hesitation or failure in fully adopting AI-driven BMI. 3️⃣ Resources and Capabilities: Successful AI-driven BMI requires extensive resources and capabilities, including a robust data strategy, skilled digital talents, adequate system infrastructure, and sufficient financial resources. These elements are essential for collecting, processing, and leveraging data to drive AI applications and business model innovations. 4️⃣ Application of AI: Implementing AI in business models involves understanding the current model, formulating an AI strategy, and selecting appropriate AI tools and technologies. Multidisciplinary teams play a crucial role in managing AI projects, ensuring effective rollout, communication, visualization, and continuous improvement of AI initiatives. 5️⃣ Implications: AI can support, enable, innovate, or disrupt business models. It enhances existing processes, redefines operations, creates new value propositions, and can lead to industry-wide transformations. The implications of AI-driven BMI are profound, offering incremental improvements, fundamental operational changes, innovative new services, and disruptive market shifts. 6️⃣ Management and Organizational Issues: Effective management is critical for driving AI initiatives and facilitating business model changes. This includes cultivating an AI-centric organizational culture, acquiring practical AI experience, rethinking governance structures, and aligning AI initiatives with company strategy. Addressing cultural deficits, fostering agility, and democratizing AI within the organization are essential for successful AI-driven BMI. ✍🏻 Philip Jorzik, Sascha P. Klein, Dominik K. Kanbach, Sascha Kraus, AI-driven business model innovation: A systematic review and research agenda, Journal of Business Research, Volume 182, 2024, 114764, ISSN 0148-2963. DOI: 10.1016/j.jbusres.2024.114764

What kind of AI-native company are you building? We talk a lot about model quality, UX, and moats. But behind the best AI startups today is something more fundamental: A business model that’s designed for how AI systems actually work in the wild. Over the last year, we’ve seen four distinct business models gain traction. They differ in interface depth, ops intensity, margin structure, and entanglement with customer reality. 👇 Dive into the full breakdown below: Here’s a quick primer: 🔹 Model 1: Product-Only Distribution compounds faster than AI models decay. These companies win by embedding into daily workflows—with UX, trust, and distribution that outlasts any one AI model. Examples: Cursor, Perplexity, MotherDuck 💡 Cursor isn't winning on model access. It's winning because it mirrors how devs context-switch, debug, and flow through large codebases. 🔹 Model 2: Product + Embedded Engineering You don’t build the spec in the lab. You build it in the field. These companies embed engineers alongside customers—not to consult, but to co-develop domain-specific systems that actually hold up. Examples: Harvey, Adaptional, CurieTech AI 💡 Harvey doesn’t sell “legal AI.” It builds copilots with Am Law firms, tuned to real workflows and risk psychology. 🔹 Model 3: Full-Stack Services: Where AI is embedded Customers aren’t buying tools. They’re buying outcomes. These companies offer AI-powered services—not software—with control over data, execution, and continuous feedback. Examples: LILT AI, Town 💡 Lilt delivers global localization as a managed service, blending human expertise with AI at every step—from content routing to tone correction. 🔹 Model 4: Roll-Up + AI Don’t start from zero. Start from ops. Infuse with AI. These companies acquire expert-heavy physical businesses (e.g. warehouses, pharmacies) and embed AI into labor, logistics, and trust loops. Examples: stealth roll-ups in logistics, healthcare, robotics 💡 A warehouse roll-up using AI to route robotic arms, triage edge cases, and compound labor—not replace it. Across all four models, one truth keeps surfacing: AI is not the product. It’s the substrate. The best companies aren’t “AI-powered tools.” They’re systems—engineered for throughput, refined in production, and impossible to unbundle. Huge thanks to Ashish Thusoo, Jordan Tigani, Suril Kantaria, Dylan Reid, Jocelyn Goldfein, and Annelies Gamble for sharing insights, counterexamples, and lived experiences.

AI is finally making services businesses scalable—and—exciting to VCs. The global services market is in the trillions of💰s, far larger than today’s software market. Yet, services businesses haven’t been the darlings of venture capital, as they were perceived to lack rapid scaling potential. 𝗔𝗜 𝗶𝘀 𝗰𝗵𝗮𝗻𝗴𝗶𝗻𝗴 𝘁𝗵𝗮𝘁. By blending AI seamlessly with human expertise, there is an opportunity to get into much larger markets with models that have the potential to scale in ways services - or even SaaS businesses - can't. For example, instead of offering a marketing SaaS, an AI-powered Service-as-Software business can deliver what the customer really wants: high-quality leads or compelling content. We’ve seen this potential firsthand through Emergent Ventures’ investments in multiple AI-powered companies that leverage humans-in-the-loop. These models resonate with B2B customers because they offer faster, clearer paths to value—reliable outcomes delivered with greater efficiency. For many customers, it’s a significant upgrade over traditional agency or service-provider relationships. While the potential is huge, only a fraction of AI-powered services startups will scale. 𝗦𝘂𝗰𝗰𝗲𝘀𝘀 𝗱𝗲𝗽𝗲𝗻𝗱𝘀 𝗼𝗻 𝘁𝗵𝗼𝘂𝗴𝗵𝘁𝗳𝘂𝗹 𝗲𝗮𝗿𝗹𝘆 𝗰𝗵𝗼𝗶𝗰𝗲𝘀 𝗮𝗻𝗱 𝗲𝘅𝗰𝗲𝗽𝘁𝗶𝗼𝗻𝗮𝗹 𝗲𝘅𝗲𝗰𝘂𝘁𝗶𝗼𝗻. Here’s what we have learned works well: 𝟭. 𝗔𝗜-𝗛𝘂𝗺𝗮𝗻 𝗦𝘆𝗻𝗲𝗿𝗴𝘆: AI and software should do the heavy lifting, with humans involved strategically— e.g. for validating AI output, edge cases, enabling adoption, or acting on AI insights. Over time, reduce human input as the AI learns, and models improve. Target 60%+ initial gross margins, with a path to SaaS-like 75%+ margins over time. 𝟮. 𝗦𝗰𝗮𝗹𝗮𝗯𝗹𝗲 𝗛𝘂𝗺𝗮𝗻 𝗜𝗻𝘃𝗼𝗹𝘃𝗲𝗺𝗲𝗻𝘁: The dependency on hiring & training humans should not constrain scale and economics. Have a path to tapping into freelancers or agency partners. Leverage human experts in a high-talent location such as India. 𝟯. 𝗥𝗲𝗰𝘂𝗿𝗿𝗶𝗻𝗴 𝗥𝗲𝘃𝗲𝗻𝘂𝗲: Focus on high-value, recurring use-cases to ensure subscription-based revenue with strong net revenue retention (NRR). 𝟰. 𝗣𝗿𝗶𝗰𝗶𝗻𝗴 𝗣𝗼𝘄𝗲𝗿: Iterate to a solution that can command higher pricing, and a model that aligns incentives with customers, e.g. based on outcomes. 𝟱. 𝗗𝗮𝘁𝗮 𝗠𝗼𝗮𝘁𝘀: Build solutions that improve with use, creating compounding competitive advantages over time. 𝟲. 𝗠𝗼𝗱𝘂𝗹𝗮𝗿 𝗧𝗲𝗰𝗵: Architect a stack that can evolve with AI advancements. 𝟳. 𝗙𝘂𝗹𝗹-𝗦𝘁𝗮𝗰𝗸 𝗧𝗲𝗮𝗺: A founding team that has the technical expertise to build and rapidly improve complex AI-powered solutions, and deep operational acumen. A rare combination. These are complex businesses to build, and the right playbooks are yet to be perfected. But where this works, 𝗦𝗲𝗿𝘃𝗶𝗰𝗲𝘀-𝗮𝘀-𝗦𝗼𝗳𝘁𝘄𝗮𝗿𝗲 𝗔𝗜 𝗯𝘂𝘀𝗶𝗻𝗲𝘀𝘀𝗲𝘀 𝘄𝗶𝗹𝗹 𝗿𝗲𝗱𝗲𝗳𝗶𝗻𝗲 𝗺𝗮𝗻𝘆 𝗕𝟮𝗕 𝗰𝗮𝘁𝗲𝗴𝗼𝗿𝗶𝗲𝘀 📈 #EnterpriseAI #startups #vc #SaaS

In the next phase, AI agents will be autonomous economic participants. The economy will evolve dramatically as agents operate continuously, share perfect information, and rapidly adapt. However our existing human-centric economy is not designed for agents. A very interesting paper “Unlocking AI Agents Potential Through Market Forces” (link in comments) explores in detail the barriers to the economic potential, and the enablers to move past those. 🚧 Human-centric infrastructure as a barrier. The current digital ecosystem was built for human users, with interfaces, identity verification, and payment systems designed around human behavior. These constraints prevent AI agents from seamlessly integrating into digital economies, limiting their ability to create and exchange value autonomously. 🔍 Challenges in service discovery. AI agents struggle to find and evaluate services because discovery mechanisms—such as industry events, peer recommendations, and human-oriented documentation—are not machine-readable. Future solutions must include structured registries, machine-friendly descriptions, and automated indexing for real-time service discovery. 🔑 Identity and authorization limitations. AI agents lack traditional identity markers like physical documents, email addresses, and human-verifiable credentials. Current authentication methods are slow and require human intervention, making them unsuitable for machine-speed operations. Cryptographic identity systems, decentralized reputation models, and dynamic access control could solve these challenges. 🌐 Software interfaces designed for humans. Digital services currently separate human-friendly visual interfaces from APIs meant for machine interactions, creating inefficiencies for AI agents. Future systems should support adaptive, machine-readable interfaces that dynamically adjust based on the consumer, whether human or AI. 💰 Payment systems block AI participation. Online transactions rely on human verification, anti-bot measures, and rigid business models like subscriptions and credit card payments. AI-friendly payment solutions should incorporate cryptographic attestation, machine-scale wallets, and real-time micropayments to enable seamless economic activity. 🚀 Future infrastructure for AI-driven markets. To fully integrate AI agents into digital markets, the ecosystem needs machine-readable service discovery, scalable identity and authorization systems, flexible payment mechanisms, and new market protocols. These advancements will unlock economic efficiency, innovation, and autonomous value creation at an unprecedented scale. This is a central theme in my work on AI-driven business model innovation, I will be sharing a lot more related insights on this.

FDEs becoming one of the most in-demand roles in tech is an early sign that business models are shifting. Labor markets often show us where value is moving first, and in AI the rise of forward-deployed teams is that signal. It shows that services are becoming a core part of how AI companies are built. In my last post Trading Margin for Moat, I argued that forward-deployed teams are one way this shows up in the go-to-market motion. What’s emerging now is that while many startups use a similar services-led approach, they’re actually solving for very different end goals. Broadly, I see three categories: 1. AI Applications: These are product-first companies that rely on forward-deployed teams to ensure early customer success. The focus is on implementation, integration, and product discovery — covering gaps while the product matures into a repeatable motion. Over time, these businesses may rely less on in-house delivery and more on an ecosystem of partners and providers. 2. AI Service Delivery: These startups look more like next-generation consulting firms. They combine LLMs and software with a high-touch services model, typically targeting budgets historically owned by Accenture or Deloitte. The motion is often less repeatable, but the upside is in unlocking higher levels of service efficiency. In many cases, they’ll evolve into ecosystem partners for AI application companies — and we’re already seeing early partnerships with model providers like OpenAI. 3. Vertically Integrated AI Services / AI Rollups: This camp is about building full-stack services businesses where AI is deeply embedded into delivery. Some are rolling up existing categories, others are building proprietary software to automate workflows and grow organically. Early entrants are already showing both margin expansion and meaningfully better customer experiences, which is why this model has quickly become one of the most popular playbooks in Silicon Valley. Markets have always been skeptical of services, but services layers are often where the real defensibility gets built. In every technology wave, the ability to weave new tools into workflows, relationships, and distribution is what turns a breakthrough into an enduring business. AI will be no different.

Is it AI vs humans or AI + humans? The answer is there isn't a one-size-fits-all approach. Instead, AI applications fall into three distinct categories: 1. Auto-pilots: Fully Autonomous Agents 🤖 These can run entire workflows without human intervention. Think 11x for handling SDR outreach or Sierra for managing support tickets. They excel at repetitive tasks where human judgment isn't critical. But let's be real - we're still in the early days here. 2. Co-pilots: Your Intelligent Assistants 🤝 These augment human capabilities rather than replace them. Think Sybill for account executives or Cursor for developers. They're not eliminating jobs; they're making professionals exponentially more productive. Applicable where human judgment and relationship-building are paramount. 3. AI-Assisted Services ⚡  This model disrupts existing services businesses with a more efficient OpEx structure, using AI to dramatically cut costs and increase margins. VCs like Andreessen Horowitz and Sequoia Capital are betting big on this model. Examples include Alma for legal, Zeni for accounting, and Immersa.ai for DataOps. For these functions, domain expertise trumps company-specific knowledge. Picking the right category is critical to success. Take enterprise sales, for instance: Could you fully automate an Account Executive's role? Not a chance. Why? Because enterprise deals require trust, relationship nuance, and understanding unspoken needs - things AI simply can't replicate. The Framework for Choosing: • If full automation is possible (like routine SDR tasks) → go for it • If relationships drive your business → augment, don't replace • If your team have deep, generalizable domain expertise → choose AI-assisted services The future isn't about replacing humans with AI. It's about understanding the Jobs To Be Done in each of three distinct paths and choosing the right one for your specific usecase. What do you think? Which other business functions do you see fitting clearly into one of these categories? #AI #Sales #Innovation #FutureOfWork #BusinessStrategy

🚀 AI-Driven Business Strategies: The 2025 Playbook Every Leader Needs The future of business isn’t just digital—it’s intelligent. By 2025, AI isn’t a “tool” in your strategy; it’s the foundation. Here’s how top companies are rewriting the rules: 🔑 Key Trends Shaping 2025 1. Hyper-Personalization at Scale       → AI deciphers customer behavior to deliver tailored experiences, boosting loyalty and revenue.     2. Decision Intelligence       → Predictive analytics + AI = automated, real-time decisions (supply chains, pricing, logistics).     3. Generative AI Revolution       → From marketing copy to synthetic data testing, GenAI slashes time-to-market by 70%+.     4. Operational Overdrive       → AI optimizes routes, inventory, and workflows—cutting costs while scaling efficiency.     5. Ethical AI & Cybersecurity       → Trust is non-negotiable. Leaders bake transparency into AI and deploy it to combat threats.     Why This Matters🌍 - $1.3T Market by 2032: Early adopters are already outpacing competitors. - Dynamic Pricing: AI shifts pricing from static to behavior-driven, maximizing margins. - Swarm Learning: Cross-department AI collaboration unlocks innovation silos.  🛠️ The Execution Gap Success isn’t about tech—it’s about strategy: ✅ Start small, iterate fast. ✅ Invest in AI fluency (teams and tools). ✅ Prioritize data quality like it’s oxygen. The Bottom Line: Companies that treat AI as a strategic partner—not a cost center—will dominate. The rest? Still debating ChatGPT prompts.

These traditional business models are becoming obsolete. (Even YC's guide needs an update for the AI era) Many founders are still building companies using outdated frameworks. Here's what they're missing: The end of "pure play" models → Anthropic charges subscriptions + usage + enterprise deals → Midjourney blends consumer subscriptions with enterprise licensing → OpenAI's hybrid model: API usage + enterprise + consumer subs The old "pick one model" playbook is dead. Speed is the new moat → Hugging Face added enterprise offerings while maintaining open-source growth in about 18 months → Stability AI launched commercial products in about a year → Traditional 3-5 year GTM timelines don't work anymore Companies that stick to old rollout schedules are being left behind. Distribution is built in → Vercel’s growth strategy creates a pathway for users to become enterprise customers, with a focus on high-potential prospects → Replicate provides tools for developers to monetize their AI models, enabling marketplace-like opportunities → The line between product and distribution is disappearing The reality? These traditional models worked in a world of slower innovation cycles. But AI is forcing a complete rethink of how value is created and captured. What actually works now: → Building multiple revenue streams from day one → Launching fast, then adding enterprise features → Using the product itself as the go-to-market strategy The fundamentals of business still matter. But the playbook for executing them has changed entirely. Curious: Are you still following traditional business models, or have you adapted your approach for the AI era? #ai #business #entrepreneurship #startups