Durable Sessions AI UX vs Scientific Business Strategy
// TL;DR
These two frameworks solve completely different problems and are never interchangeable. If you're building or fixing an AI-powered product whose chat streaming breaks on disconnect, can't work across devices, or lacks user control, use the Christensen Durable Sessions AI UX Framework. If your business is stuck, plateaued, or scaling without a coherent strategy, use the Steve Patrick Scientific Business Strategy Framework. Pick based on whether your problem is technical product architecture or business direction.
// HOW DO THEY COMPARE?
| Dimension | Christensen Durable Sessions AI UX Framework | Steve Patrick Scientific Business Strategy Framework |
|---|---|---|
| Best for | Engineering and product teams building AI chat or agent-driven products with streaming UX issues | Business owners, founders, and executive teams needing a coherent growth or turnaround strategy |
| Problem domain | Technical: real-time streaming architecture, connectivity resilience, multi-device UX | Strategic: identifying the root business constraint and building a tested plan to overcome it |
| Complexity | High — requires understanding of SSE, WebSockets, pub/sub, and distributed systems concepts | Moderate — conceptually accessible but demands rigorous honesty and iterative thinking |
| Time to apply | Days to weeks for architecture redesign; hours for an audit | 1–3 day intensive strategy session, plus ongoing iteration as constraints shift |
| Prerequisites | An existing AI product with a streaming architecture; engineering team capable of infrastructure changes | A business with revenue or market presence; willingness to confront root-cause problems honestly |
| Output type | Architecture redesign: a Durable Sessions layer, transport upgrade plan, and validated resilience checklist | A single Strategic Problem definition, ICP, competitive posture, and DPE-filtered strategic option ready for testing |
| Creator background | Mike Christensen (Ably) — real-time infrastructure and AI UX engineering | Steve Patrick — business strategy consulting and competitive positioning |
| Framework type | Technical diagnostic and architectural pattern | Business strategy methodology rooted in scientific method and Porter's competitive theory |
| Iterative nature | Applied once per product architecture; revisited when adding new surfaces or agents | Explicitly iterative — each solved constraint reveals the next; designed as a repeating loop |
| Key risk if skipped | AI product feels like a fragile demo: streams break, no multi-device sync, no stop button | Team wastes months executing on the wrong problem, building org charts and priority lists without strategic foundation |
What does the Christensen Durable Sessions AI UX Framework do?
The Christensen Durable Sessions AI UX Framework, created by Mike Christensen of Ably, diagnoses why AI chat and agent-driven product experiences break under real-world conditions — and provides a concrete architectural fix. It identifies the Single-Connection Trap: the default pattern where streaming an AI response over a direct HTTP/SSE connection means the stream dies if the connection drops.
The framework evaluates your product against three foundational capabilities: Resilient Delivery (streams survive disconnections), Continuity Across Surfaces (the session follows users across tabs and devices), and Live Control (users can steer, interrupt, or cancel the agent mid-generation). The solution is a Durable Sessions layer — a persistent, shared resource between agent and client — implemented via pub/sub channels. Agents write events to the session; clients subscribe. Neither holds a direct pipe to the other.
This framework is essential for any team whose AI product works in demos but breaks when users switch networks, open a second tab, or press a stop button.
What does the Steve Patrick Scientific Business Strategy Framework do?
The Steve Patrick Scientific Business Strategy Framework helps business leaders diagnose the single root-cause constraint blocking growth and build a tested, multi-option strategy to overcome it. It explicitly rejects jumping into org charts, priority lists, or annual planning before the real problem is identified.
The process starts with the Five Whys drill to surface the true Strategic Problem — the one wall between the business and its Dream Outcome. A single Ideal Customer Profile (ICP) is placed at the center of every decision. The framework then walks through market positioning, competitive behavior (using Porter's three generic strategies), and resource/return modeling. Multiple strategic options are generated and filtered through the DPE test (Desirable, Practical, Economical). The winning option is tested in the real world before full commitment, honoring the distinction between Intended Strategy and Emergent Strategy.
This framework is for founders, executives, and strategy teams who suspect they're busy solving the wrong problem.
How do they compare?
These frameworks operate in entirely different domains and solve different classes of problems. The Durable Sessions framework is a technical architecture pattern for product engineering teams. The Scientific Business Strategy framework is a strategic planning methodology for business leaders. There is virtually no overlap in audience, inputs, outputs, or application context.
The Durable Sessions framework assumes you already have a product and an AI-powered experience that needs to be made resilient and production-grade. It requires engineering knowledge of streaming protocols, real-time infrastructure, and distributed systems.
The Scientific Business Strategy framework assumes you have a business that needs strategic direction. It requires honest introspection, competitive analysis, and financial modeling. It is protocol-agnostic and has nothing to do with software architecture.
Where they share conceptual DNA is in their insistence on diagnosing the real problem first. Christensen warns against bolting on Redis buffering and retry logic without addressing the architectural root cause. Patrick warns against building org charts and priority lists without naming the Strategic Problem. Both frameworks punish teams that skip diagnosis and jump to action.
Which should you choose?
The answer depends entirely on what problem you're facing:
Choose the Christensen Durable Sessions AI UX Framework if:
- You are building or maintaining an AI product with streaming chat or agent responses.
- Users lose responses on disconnect, can't see activity on a second device, or your stop button doesn't work properly.
- Your engineering team is drowning in retry, resume, and relay logic inside agent code.
- You need a concrete architectural pattern — not business advice.
Choose the Steve Patrick Scientific Business Strategy Framework if:
- Your business is stuck, plateaued, or growing without coherent direction.
- Your team is tempted to jump into tactics (hiring, marketing campaigns, new products) without first identifying the root constraint.
- You need to decide where to compete, how to compete, and whether your plan is financially viable.
- You need strategic clarity — not infrastructure guidance.
If you are an AI product company facing both a broken streaming UX and unclear business strategy, apply them sequentially: strategy first (Patrick) to confirm you're solving the right market problem, then architecture (Christensen) to ensure the product experience is production-grade. They complement each other but never substitute for each other.
// FREQUENTLY ASKED QUESTIONS
Can I use the Durable Sessions framework for business strategy?
No. The Durable Sessions framework is a technical architecture pattern for AI product streaming infrastructure. It does not address market positioning, competitive strategy, customer selection, or financial planning. For business strategy, use the Steve Patrick Scientific Business Strategy Framework.
Do I need engineering skills to use the Steve Patrick strategy framework?
No. The Scientific Business Strategy Framework is designed for business leaders, founders, and executive teams. It requires honest problem diagnosis, competitive thinking, and basic financial modeling — but no software engineering knowledge. It is entirely non-technical.
What is a Durable Session in the Christensen framework?
A Durable Session is a persistent, shared resource sitting between the AI agent layer and the client layer. Agents publish events to it; clients subscribe. It survives disconnections, supports multiple devices, and enables live control — solving the three foundational problems of AI product streaming UX.
What is the DPE filter in the Steve Patrick framework?
DPE stands for Desirable, Practical, and Economical. Every strategic option must pass all three gates: the team must genuinely want to pursue it (D), have or be able to acquire the capabilities to execute it (P), and the financial model must show viable cash flow and return on invested capital (E).
Which framework should an AI startup use first?
Start with the Steve Patrick Scientific Business Strategy Framework to confirm your ICP, market positioning, and competitive behavior. Once you have strategic clarity, apply the Christensen Durable Sessions framework to ensure your AI product's streaming architecture is resilient, multi-surface, and supports live user control.
Does the Durable Sessions framework replace the Vercel AI SDK or LangChain?
No. It sits alongside them. The framework adds a session layer between your existing agent tooling (Vercel AI SDK, LangChain, etc.) and the client. Your agents still generate responses using these tools; the Durable Sessions layer handles delivery, resilience, and multi-surface sync.
Can these two frameworks be used together?
Yes, but they solve different problems and are applied independently. Use the Patrick framework to set business direction and validate your strategic focus. Use the Christensen framework to fix the technical delivery layer of your AI product. They are complementary, not competing.
What is the biggest mistake each framework warns against?
The Durable Sessions framework warns against building resume and retry logic inside the agent code itself — that complexity belongs in the session layer. The Scientific Business Strategy framework warns against skipping the Strategic Problem definition and jumping straight to org charts and priority lists.