Porter's Five Forces for PM interviews

Porter’s Five Forces maps the structural forces that shape profitability in a market: competitive rivalry, threat of new entrants, threat of substitutes, bargaining power of buyers, and bargaining power of suppliers. Published by Michael Porter in HBR in 1979. In a PM interview, the failure mode is walking through all five in order with generic examples. The pass is identifying which two or three forces are dominant and closing with a recommendation.

When to reach for it

Five Forces earns its place on market entry questions, competitive response questions, and build-vs-buy-partner questions where the answer depends on who holds structural power. It is weak for single-product-within-large-company questions: if the question is about prioritizing features inside an existing product, reach for a prioritization framework instead.

Do not name the framework unless you are about to run it selectively. “I’d run a Five Forces analysis” followed by equal treatment of all five forces reads as memorization.

The five forces, with PM-specific language

Competitive rivalry: intensity of direct competition across number of rivals, differentiation, and switching costs. If rivalry is intense and differentiation is low, pricing and retention strategy matters more than the feature roadmap.

Threat of new entrants: asymmetric in AI markets. Open-source models (Llama, Mistral, DeepSeek) collapsed the model-layer barrier; DeepSeek trained a frontier-competitive model on roughly 2,000 Nvidia H800 GPUs at a fraction of prior incumbent cost. The real entry barriers are now proprietary data, embedded workflow, and distribution. “AI lowers barriers to entry” without that distinction will not clear the bar.

Threat of substitutes: the most important force in 2026 and the most systematically underused. Substitutes are not named competitors; they are adjacent tools that absorb your use case. Cursor made the IDE the product and displaced standalone code review tools. Notion AI displaced standalone writing assistants by being already open in the tab. Run this analysis against every adjacent tool in the user’s workflow, not just rivals in your category.

Bargaining power of buyers: shifted wherever AI agents mediate purchasing. A procurement bot or Perplexity query compares products in seconds; brand loyalty and switching-cost inertia erode. High buyer power means functional stickiness matters more than marketed differentiation. Tock beat OpenTable by recognizing this: restaurants had leverage against per-cover fees, and flat SaaS pricing converted suppliers into advocates.

Bargaining power of suppliers: dependency on cloud infrastructure and foundation model providers (OpenAI, Anthropic, Google). If your product is built on a single provider’s API and that provider enters your product category, supplier power becomes existential. The OpenAI-Microsoft and Google-Workspace partnerships show how supplier consolidation redrew competitive maps faster than organic rivalry.

Which force to lead with

The prompt type tells you where to start:

• Market entry: Lead with substitutes (what already solves this job?) and entry threat (model access or data, which is the real barrier?).
• Defend a product: Lead with rivalry and substitutes (what adjacent stack can absorb our use case?).
• Build vs. buy vs. partner: Lead with supplier power and entry threat (who controls the input, and at what cost can we replicate it?).

State the dominant one or two forces upfront. Skip or briefly name low-relevance ones explicitly. Close with a recommendation tied to the analysis, not a summary of the forces.

Worked example: entering the enterprise AI assistant market

“The most important force here is substitutes from adjacent distribution. Microsoft Copilot is embedded in Office 365; Google Gemini is in Workspace. A standalone entrant competes against tools users already pay for and have open all day. Buyer power is high: IT procurement now uses AI evaluation tools that compress feature comparison to minutes, so differentiation has to be functional, not marketed. Supplier power is low at the model layer; any team can fine-tune Llama or Mistral. The real entry barrier is proprietary workflow data. I’d recommend entering via a vertical where your training data is proprietary rather than competing horizontally where the substitutes are entrenched, or partner with a distribution platform to convert that barrier into an asset.”

That answer names dominant forces, explains the mechanism, and closes with a build-buy-partner recommendation.

What weak looks like in the room

Walking through all five forces in textbook order with generic examples and no synthesis. The interviewer gets a list, not a diagnosis: forces named, no dominant force identified, no strategic implication stated, no recommendation. In 2026, weak answers also ignore AI-era dynamics or bolt on “AI is a substitute” as a throwaway line without explaining the mechanism.

Related: SWOT for PMs, Beyond frameworks, Respond to a competitor launching your feature