Lean startups (Blank and Ries)

Lean Startup treats a new venture as a search for a sustainable business model under uncertainty. Eric Ries combined Steve Blank’s customer-development method with iterative product engineering in The Lean Startup (2011), turning assumptions into experiments before the organisation scales them.

When to use it

• Use the approach when customer, problem, solution, channel or business-model uncertainty is high. It applies to independent ventures and new initiatives inside established organisations, with additional governance where failure could harm people or violate regulation.

Origins

Peter Drucker wrote in 1954 that business exists to create a customer. Steve Blank operationalised that emphasis through customer development, while Eric Ries integrated it with agile engineering and lean ideas.

Blank’s intellectual starting point repeats Drucker’s 1954 customer focus: a startup should test whom it serves and why before treating a plan as fact.

What it is

Blank’s The Four Steps to the Epiphany (2003) described a search process for a repeatable, scalable model. The Startup Owner’s Manual (2012), written with Bob Dorf, expanded the method. Blank’s later article in 2013 helped establish three connected principles:

• Business-model hypotheses: use a canvas to make assumptions about value, customers, channels, revenue, costs, partners, activities and resources visible.
• Customer development: engage with prospective customers and other stakeholders to test those assumptions.
• Agile engineering: build incrementally so product decisions respond to evidence.

A business plan can coordinate execution once important assumptions are supported. Before then, false precision can conceal uncertainty; the canvas is a testable model rather than a smaller plan.

How to use it

Translate the idea into hypotheses and rank them by importance and uncertainty. Identify the smallest ethical test that can produce decision-relevant evidence. “Get out of the building” means observing and interviewing people in context, not pitching until someone agrees.

Customer development moves through discovery, validation, customer creation and company building. Discovery tests the problem and proposed value. Validation looks for behavioural evidence such as use, commitment or payment. Weak evidence may trigger iteration, a pivot or termination. Only after repeated evidence should the venture invest heavily in demand generation and formal scaling.

SEARCH EXECUTION

Build–Measure–Learn provides the engineering loop. Build a minimum viable product (MVP) that tests a specific assumption; define the measure before release; learn whether evidence supports the hypothesis and which decision follows.

“Minimum” does not excuse unsafe, inaccessible or deceptive products. In health, finance, employment, infrastructure and other high-impact settings, prototypes require risk controls, informed participation and relevant expert or regulatory review. A landing page, concierge service, mock-up or technical spike may test the assumption without exposing users to an immature system.

Use qualitative evidence to understand why and quantitative evidence to estimate behaviour. Avoid vanity metrics and post-hoc success criteria. Record experiment design, sample, result, limitations and decision.

A pivot changes a fundamental hypothesis while retaining validated learning. Repeated pivots without a clear falsification rule can become avoidance. Stop when the opportunity no longer justifies further cost, risk or attention.

Top practical tip

Write the assumption and decision before building the MVP: “If we observe this evidence, we will continue; otherwise we will change or stop.”

Top pitfall

Do not use experimentation to transfer risk to customers or evade planning. Learn Startup reduces uncertainty; it does not remove legal, ethical, safety or operational obligations.

Further reading

Ries, E. The Lean Startup.

Blank, S. The Four Steps to the Epiphany.