Objective: Learn the AI product taxonomy framework for categorizing AI products and features by their strategic role and differentiation potential.
"Not all AI features are created equal. Some are table stakes; others are transformative. Knowing the difference shapes your entire product strategy."
AI Product Strategy Handbook
If your AI feature is "smart sorting" in email, congratulations: you've built table stakes. Your competitor has the same thing. You're not differentiated, you're just not fired.
6.1 AI Product Taxonomy
AI products and features fall into distinct strategic categories. Understanding this taxonomy helps product teams make build/buy/bake decisions, allocate resources effectively, and avoid common strategic mistakes.
The Four Categories of AI Products
Category 1: AI as Table Stakes
AI capabilities that users expect to be present. Not differentiating, but omitting them is costly. Examples include basic search relevance, spam filtering, and recommendation feeds.
Category 2: AI as Feature Differentiator
AI capabilities that differentiate your product from competitors when done well. These create competitive advantage through superior performance or unique functionality.
Category 3: AI as Core Product
AI is the primary value proposition. The entire product experience centers on AI capability. Examples include AI writing assistants, code generation tools, and AI research assistants.
Category 4: AI as Platform
AI infrastructure that enables other products or features. These create ecosystem effects and can become strategic moats.
Table Stakes AI
Table stakes features are expected baseline capabilities. Users assume they exist and do not typically make purchasing decisions based on them.
Characteristics
Table stakes features share several key characteristics. They offer low differentiation because users expect them and they do not create preference between competing products. They carry high implementation cost, meaning they must be built but provide minimal competitive moat once implemented. They represent competitive necessity, where omitting them is a significant disadvantage that would put the product at a substantial deficit compared to alternatives.
Strategic Approach
Build or buy these efficiently. Do not over-invest. Use commodity AI services rather than custom development. Focus energy on features that differentiate.
Common table stakes AI features include spam filtering in email products, which users expect to work without them thinking about it, search relevance in content platforms that helps users find what they need, fraud detection in financial products that protects users from malicious activity, autocomplete in text editors that speeds up writing, and recommendation feeds in content apps that surface relevant material. These capabilities are expected rather than differentiated, but still necessary for competitive parity.
Some teams conclude that since table stakes features are not differentiating, they should deprioritize them entirely. This is dangerous. Table stakes done poorly create user churn even if competitors do them well. Your AI search returning irrelevant results when competitors return relevant ones is not neutral; it actively drives users away. Invest enough in table stakes to be competitive, then focus disproportionate investment on differentiators.
Feature Differentiators
These AI features distinguish your product from competitors when executed well.
Characteristics
Feature differentiators possess distinct characteristics that make them worth strategic investment. They have high differentiation potential, meaning they can create meaningful competitive advantage when executed well. They require variable investment, typically demanding significant resources to achieve excellence rather than mere adequacy. They have user-visible impact, directly affecting user experience and outcomes in ways that users notice and value.
Strategic Approach
Invest heavily in these. Build proprietary data advantages, fine-tune models for your specific domain, and design unique workflows that competitors cannot easily replicate.
Feature differentiators in practice include superior code completion in a developer tool that makes developers dramatically more productive, unique document understanding in a legal tech product that surfaces relevant precedents others miss, personalized learning paths in an EdTech platform that adapt to each student's strengths and weaknesses, and catchy copy generation in a marketing tool that produces compelling content better than competitors. Differentiation requires sustained investment in capabilities that competitors cannot easily replicate.
Differentiation requires sustained investment in capabilities that competitors cannot easily replicate.
Core AI Products
These products exist primarily because of AI capability. The entire product experience centers on what AI enables.
Characteristics
Core AI products have fundamental characteristics that set them apart from other categories. In these products, AI is the product itself, meaning that if you remove the AI capability, nothing meaningful remains of the product experience. The stakes are high because AI quality directly determines product success in ways that do not apply to supplementary features. These products require full-stack investment, demanding deep investment in AI infrastructure including models, evaluation systems, and reliability engineering.
Strategic Approach
These require the most investment but also offer the highest returns if successful. Build or significantly customize foundation models. Invest in evaluation infrastructure. Design for AI reliability from the start.
Core AI products in the market include AI writing assistants like Claude and ChatGPT that help users generate and refine text, AI code generation tools like GitHub Copilot that translate natural language instructions into working code, AI image generation systems like Midjourney and DALL-E that create visual content from text descriptions, and AI research tools like Perplexity and consensus.ai that help users find and synthesize information from across the internet.
AI Platforms
AI infrastructure that enables other products and features, often creating ecosystem effects.
Characteristics
AI platforms have distinct characteristics that create unique strategic value. They enable others to build on top of them, serving as infrastructure that other products and features depend upon. They generate network effects where more users create more data, which in turn leads to better models and more attractive capabilities for new users. They exhibit platform economics that can become strategic moats, as the combination of network effects and switching costs creates durable competitive advantages that are difficult for competitors to replicate.
Strategic Approach
These require significant investment but create durable advantages. Build for adoption. Create developer ecosystems. Focus on enabling third-party innovation.
AI platforms in the current market include the OpenAI API that enables countless applications to access powerful language model capabilities, the Anthropic Claude API that provides access to frontier model capabilities for developers, AI model marketplaces that allow developers to discover and deploy specialized models, and AI fine-tuning platforms that enable organizations to customize foundation models for their specific domains.
Making Taxonomy Decisions
For each AI feature you are considering, work through these four questions systematically. First, determine what category it falls into, whether table stakes, differentiator, core, or platform, based on its strategic role and differentiation potential. Second, ask whether this is the right category, considering whether it should perhaps be a differentiator rather than table stakes, or a core capability rather than a differentiator, depending on your product strategy. Third, determine what investment level is appropriate, matching your investment to the category rather than over-investing in table stakes or under-investing in differentiators. Fourth, clarify what the competitive positioning is, understanding how this feature creates or erodes advantage in the market.
Before committing to a taxonomy category, define how you will measure whether the categorization was correct. A micro-eval for AI taxonomy tests: user perception of value by category, competitive differentiation by category, and ROI by category over time. Stripe's eval-first insight: they initially categorized code search as "table stakes" but their eval showed it drove 30% of developer retention. After re-categorizing as "core differentiator," they invested appropriately and saw NPS jump 25 points.
What's Next?
Next, we explore Build vs. Buy vs. Bake Decisions, understanding how to make strategic choices about AI development versus acquisition.