Richard Rumelt on Good Strategy Bad Strategy

Richard Rumelt in conversation with Lenny Rachitsky on Lenny’s Podcast. Rumelt covers the core elements of strategy, the anatomy of bad strategy, the role of power and asymmetry, organisational barriers to execution, and how startups should think about strategic bets.

Key ideas

• Strategy is problem-solving, not goal-setting. A strategy is a design for overcoming a high-stakes challenge. The three essential elements — the kernel — are: (1) a diagnosis of the situation, (2) a guiding policy for how to address it, and (3) coherent actions that implement the policy without contradicting each other. Any document missing one of the three is not a strategy.
• Bad strategy mistakes ambitions for diagnosis. Listing goals, inserting fluff, or producing mission/vision documents are the most common failure modes. A real diagnosis names why the problem exists, not merely that a gap exists.
• Call it an action agenda, not a strategy. Rumelt advises stripping the word “strategy” and asking: what are we actually going to do about this specific challenge? Prioritisation forces focus; focus is the fundamental source of competitive power.
• Power comes from asymmetry. Competitive advantage is always rooted in some asymmetry — a unique skill, relationship, reputation, user base, or network effect — that shifts the odds. Power must be focused on a target that can actually be affected; diffusing effort across many targets dissipates it.
• The crux is the hardest part of the problem. Named after the hardest move in a rock climb, the crux is the challenge whose resolution unlocks progress on everything else. Skilled designers and strategists focus on this difficulty first; insight emerges from immersion in the problem, not from picking a strategy off a list.

The kernel

Rumelt’s three-element model for any valid strategy:

On power and asymmetry

Sources of power Rumelt discusses:

• Network effects — user base creates value that grows with scale; dominant in social media, platforms, and generative AI data advantages.
• Reputation and relationships — IBM’s Gerstner leveraged enterprise relationships as the primary asset post-mainframe.
• Proprietary knowledge / skill — being first to recognise something, or attracting talent by being at the cutting edge.
• Specialisation — the antidote to being beaten by a giant; owning a niche where the market leader’s scale is irrelevant.

On bad strategy

Four markers:

1. Fluff — vague, abstract language that sounds strategic but says nothing.
2. Failure to diagnose — jumping straight from problem statement to therefore without the because.
3. Goals masquerading as strategy — profit targets and market-share aspirations are ambitions, not strategies.
4. Incoherent actions — goals that contradict each other (e.g. grow faster and increase return on equity simultaneously).

On organisational dynamics

Complex organisations resist focused strategy because:

• People have private interests that their stated positions conceal.
• Every piece of information inside a large organisation comes with an agenda.
• Changing strategy means changing power relationships — who is the alpha, who controls budget.
• Modern management culture has displaced deciding with leading, creating hesitancy to make hard choices.

Nokia case: replacing engineers with lawyers and accountants, combined with a matrix organisation that diffused authority, meant no one could act on the CEO’s repeated calls to build a smartphone.

On startups

Startups make bets under uncertainty. Key implications:

• Clarity about the nature of the bet matters more than precision of the forecast.
• Research on Silicon Valley startups shows initial product-market hypotheses usually fail; survivors switch — different customer, different product — until something clicks.
• Requires holding two things at once: conviction that you will win and willingness to pivot when data contradicts the bet.

The foundry process

Rumelt’s consulting method for generating an action agenda inside an organisation:

1. Surface 25–50 problems on Post-it notes — making the full set visible forces participants to accept they cannot pursue everything.
2. Filter for importance (closeness to ambition) and addressability (can we actually do something?).
3. Identify one or two crux challenges at the intersection.
4. Produce an action agenda: specific things we will do, not goals we hope to achieve.

See also

• Richard Rumelt
• Hamilton Helmer on 7 Powers
• Alex Komoroske on Strategy and Complexity
• Gibson Biddle on Product Strategy
• 7 Powers