Lean Startup
Lean Startup
What Is Lean Startup?
- Test and learn fast with an MVP
- The Build-Measure-Learn loop
- If it doesn’t fit, pivot
Adopting Lean Startup changes the unit of decision-making inside a company. Where the starting point used to be approving a 'business plan,' it now becomes picking the single riskiest assumption and testing it with an experiment. So the questions in new-business meetings change too. Instead of 'How big is the market?' they become 'What did we learn from the last two weeks of experiments, and what will we check in the next two?' Budgets change as well: rather than burning KRW 500 million (≈ US$374,100) at once, you release funds in chunks of KRW 20 (≈ US$1)–30 million and decide on the next round. Reporting progress in facts confirmed, not money spent, is the heart of this methodology.
An MVP is not a 'cheaply made product' but 'a device that tests a hypothesis at the lowest possible cost.' Collecting only pre-signups through a landing page, a concierge approach where people respond by hand instead of automation, and a setup where the back end is Excel and only the front end looks like a service are all MVPs. The criteria for judging differ too. You look at cohort repurchase and conversion rates instead of vanity metrics such as cumulative visitors or downloads. There is also a boundary with Agile. If Agile is a method for 'how to build what is already decided, faster,' Lean Startup is the stage that first asks 'do we need to build this at all?'
Where it most often goes wrong in practice is mistaking an MVP for a 'half-baked, unfinished product.' In B2B, shipping lower quality isn't an experiment, it's an incident. The moment an existing customer asks, 'What's going on with this company lately?', the trust you've lost outweighs the learning you've gained. That's why you should limit an experiment to a small group of new customers or a separate brand. Conversely, if you use pivoting as an excuse to change direction every three months, no hypothesis ever gets tested to the end. Executives telling teams to experiment while judging them on first-year revenue is another common contradiction. When that happens, the person in charge stops trying to learn and picks only experiments that will produce good-looking numbers. Agree first on criteria that count failed experiments as results.
John KimAuthor of B2B Marketing Design and CEO of Nanum Consulting
Eunpyo HongExpert in company analysis, startups, and venture fundraisingIntroduced by Eric Ries in his 2011 book The Lean Startup, the approach centers on the MVP and the Build-Measure-Learn loop.
Consider a hypothetical case. Company A, an industrial supplies distributor with 90 employees and KRW 21 billion (≈ US$15,711,500) in annual revenue, planned a new business: a 'field maintenance subscription' that dispatches maintenance technicians. The initial plan was to spend 6 months and KRW 400 million (≈ US$299,300) building its own dispatch system. But the CEO put development on hold, saying, 'Let's first check whether customers will actually pay a monthly subscription fee.'
- Instead of a KRW 400 million (≈ US$299,300) system, the new business team lead set up only an Excel sheet and a KakaoTalk channel, and narrowed the riskiest assumption down to this: 'Some customers will sign a recurring contract at KRW 790,000 (≈ US$590) a month.'
- A Sales Team 2 manager picked 12 of the 30 existing accounts and made the offer; within 4 weeks, 5 had signed. The 7 that declined answered, 'Stock shortages are more urgent than maintenance.'
- Dispatch ran without a system: every morning the team lead assigned the 3 technicians by hand. Running it that way for 8 weeks produced the first hard number: an average handling time of 47 minutes per request.
- The 5 contracted accounts made only 1.8 service calls per month. After finance review, KRW 790,000 (≈ US$590) turned out to be too high, so the team restructured the pricing as a KRW 390,000 (≈ US$290) base fee plus a charge for each call beyond the included limit.
- The finance director judged, 'With these numbers, we can cut the development scope in half,' and reduced the development budget from KRW 400 million (≈ US$299,300) to KRW 170 million (≈ US$127,200). Twelve weeks later, the number of contracted accounts had grown from 5 to 23.