Frameworks that endure disruption are rare. Brian Balfour’s original Four Fits framework has long been a foundational lens for growth strategy. He recently released The Four Fits: A Growth Framework for the AI Era to capture how AI is shifting the constraints inside each dimension.
The Four Fits have always been about scaling companies to $100M+ revenue at venture speed. To succeed, all four fits must align simultaneously. In this article, I explore the evolution of the Four Fits, show what’s different in the AI era, and highlight practical implications for product leaders.
The Original Four Fits: A Baseline for Growth
Balfour’s earlier framework defined growth as the alignment of four “fits.” Each represented a crucial checkpoint for a product to scale sustainably:
-
Product–Market Fit: building something customers desperately want.
-
Product–Channel Fit: ensuring the product is built for the channels customers use to discover it. Channels don’t adapt to products; products adapt to channels.
-
Channel–Model Fit: confirming that the business model can support the chosen distribution channel.
-
Model–Market Fit: validating that the pricing or revenue model matches how customers want to buy and pay.
This model became a classic because it balanced product, distribution, and monetization in a way that founders and product leaders could repeatedly reference as companies grew. It was timeless, designed for digital products and SaaS businesses where constraints were relatively stable.
What the AI Era Changes
The AI era doesn’t replace the Four Fits. It makes each one more complex. The structure remains the same, but the definition of fit inside each category has shifted.
Product–Market Fit: Expanded Problem Spaces, Higher Expectations
AI expands the solution space dramatically. Products like GitHub Copilot or Notion AI illustrate how generative models unlock new workflows that were previously impossible. At the same time, expectations for quality are higher. Users now demand accuracy, reliability, and transparency in addition to usability. A minor hallucination in an AI tool in a critical setting isn’t just a small bug; it can undermine trust entirely.
AI can also expand or contract markets. Canva demonstrates expansion, as AI democratizes design skills and opens new creative use cases. In contrast, Chegg’s collapse highlights contraction, as ChatGPT offered instant, free homework help that eroded its core value proposition almost overnight. Market size and demand are no longer static—they shift with the pace of AI adoption.
Product–Channel Fit: Distribution is in Flux
AI is reshaping traditional distribution paths. Search is no longer a predictable acquisition channel as AI-powered answers replace static results. Social feeds increasingly surface AI-generated content, raising noise levels and reducing organic visibility. Businesses over-reliant on SEO face existential risk if traffic declines or user habits shift to new discovery environments like ChatGPT or other AI assistants. (See: Is AEO the Next SEO? )
Successful AI-native products often spread differently—through embedding in existing workflows, integrations, or viral usage loops. Consider how ChatGPT’s plugins or API usage drive adoption through developers and third-party builders. Here, the principle that channels don’t adapt to products is even more urgent. AI products must be designed to fit into emerging distribution patterns, not force new ones.
Channel–Model Fit: Cost Structures are Different
Classic SaaS benefited from near-zero marginal costs. AI products don’t. Inference costs, compute, and API expenses mean that each additional user carries significant cost weight. For example, video generation platforms have to price carefully to avoid unprofitable usage spikes. This forces experimentation with freemium, usage-based, or tiered pricing, making channel choice tightly linked to financial sustainability.
Model–Market Fit: Willingness to Pay is Unsettled
Market norms for AI pricing are immature. Some users expect AI copilots to be “free,” bundled into existing tools, while others are willing to pay a premium for specialized capabilities. Microsoft embedding Copilot into Office at a high price point shows one extreme, while free consumer-facing AI chat apps show the other.
Traditional pricing models like freemium and subscription are under stress. Token-based usage, credits, or hybrid approaches are increasingly necessary to reconcile high costs with unpredictable user value perception. For product leaders, this means willingness to pay is volatile, and pricing experiments must happen earlier and more often.
The Four Fits Side by Side
| Dimension | Original Version | AI Era Version |
|---|---|---|
| Product–Market Fit | Solve a real problem for market | Expand into new AI-native problem spaces + manage trust, safety, latency; markets can expand or collapse rapidly |
| Product–Channel Fit | Align with scalable channels like SEO, virality | New distribution paths, embedded use cases, onboarding friction; products adapt to channels, not vice versa |
| Channel–Model Fit | Unit economics that support channel | Higher compute/inference costs, usage pricing, margin pressure |
| Model–Market Fit | Pricing aligns with market norms | Fluid pricing norms, freemium/subscription challenged by token usage; willingness to pay unsettled |
Why the Framework Still Matters
It would be easy to assume that AI requires an entirely new growth framework. But Balfour’s decision to update, rather than discard, the Four Fits is telling. Growth fundamentals still apply. The difference is that AI alters the constraints inside each fit.
For example:
-
Fit is less stable. A distribution channel that looks promising today may be disrupted tomorrow by an AI platform shift.
-
Unit economics need closer monitoring. Unlike SaaS, gross margins for AI can fluctuate with usage patterns and model costs.
-
User expectations are dynamic. As models improve, what delighted users last year may feel inadequate this year.
Key Takeaways for Product Leaders
-
Use the Four Fits as a diagnostic, not a checklist. Each fit now changes faster in AI markets, requiring ongoing review, not one-time validation.
-
Integrate trust as part of product–market fit. Reliability and explainability are no longer nice-to-haves; they define fit.
-
Experiment aggressively with pricing and channels. Freemium and subscription may not hold—usage-based and hybrid models must be tested early.
-
Watch distribution dependencies. AI platform shifts (e.g., model access, API terms, or search engine redesigns) can change your growth trajectory overnight.
Conclusion
Brian Balfour’s Four Fits framework endures because it focuses on alignment across product, distribution, and monetization. The AI era doesn’t replace it but redefines what alignment looks like. For product leaders, the pressing question is not just “Do you have fit?” but “Do you have fit under AI’s shifting expectations, economics, and distribution?”
The Four Fits remain a durable growth lens—one that product leaders should now treat as a dynamic system, constantly recalibrated in response to AI-driven change.