Growth Hacking
Growth Hacking
What is growth hacking?
- Find growth levers through rapid experiments
- Works across product, data, and marketing
- Focus on the North Star metric, not vanity metrics
Once growth hacking takes hold, the first thing that changes is the question asked in meetings. It shifts from 'How much did we spend on ads this month, and how many leads came in?' to 'How many experiments did we run over the past two weeks, and how many won?' What the marketing team can touch changes too. It no longer stops at banner copy and keyword bidding; sign-up form fields, the first onboarding screen, the plan structure on the pricing table, and features like employee invitations, all of which need development resources, become experiment targets. That is why an organization where marketers, developers, and data owners work as one team is effectively a prerequisite. Without it, if only the marketing team pushes for growth, half the experiment ideas sit in the dev backlog.
In B2B, B2C-style A/B testing doesn’t work as is. A company with 3,000 monthly visitors and around 100 free-trial sign-ups needs months before the conversion gap between two groups becomes meaningful. So in Korean B2B practice, teams often compare results sequentially across separate periods instead of splitting traffic at the same time, and sharpen their hypotheses first through qualitative interviews. The line between this and performance marketing is also clear. Performance marketing raises revenue per ad dollar within set channels, while growth hacking makes the product itself part of the experiment. Once experiment ideas pile up, scoring them on impact, confidence, and ease and ranking them keeps 'the idea an executive fell in love with' from jumping the queue.
How fast experiments run ultimately comes down to the approval structure. At a company where changing a single line of landing-page copy needs sign-off from both the team lead and the division head, even the best hypothesis can only be tested once a month. It is better to open up the authority for owners to launch experiments right away within a pre-agreed scope. The same goes for record-keeping. If you share only the winning experiments and delete the losing ones, then after the owner changes, a hypothesis that has already failed gets tested again from scratch six months later, over and over.
Where practice most often goes wrong is reading growth hacking as 'a trick for growing without spending money.' If you keep repeating one-off attempts—extending the trial period, throwing up a popup, bolting on a tool—then even when an experiment wins, you can’t explain why, and it doesn’t lead to the next experiment. Premature verdicts are also a problem. Declare a winner on three days of data, roll it out everywhere, and when the metrics drift back a month later, the team concludes that 'experiments are pointless.' Fixing the minimum sample size and observation period before the experiment starts is the only safeguard. And if you push conversion rate alone and pull in customers who aren’t a fit, the cost comes back three months later as churn.
John KimInstructor John Kim, author of B2B Marketing Design and CEO of Nanum Consulting
Junyoung NohConsumer and generational trend analyst & AI content marketing specialistThe term was popularized in 2010 when marketer Sean Ellis used the phrase 'growth hacker.' It originated in the data- and experiment-driven approach to growth at Silicon Valley startups.
Consider a hypothetical case. Company A sells attendance-management SaaS and has 40 employees and annual revenue of KRW 6 billion (≈ US$4,489,000). It gets 180 free-trial sign-ups a month, yet its paid conversion rate had been stuck at 9% for over a year. It raised ad spend to KRW 20 million (≈ US$15,000) a month, but only sign-ups grew; revenue stayed flat. The marketing lead and the product owner decided to run a six-month experiment cycle.
- At the first meeting, the marketing team lead announced, 'We will stop looking at visitor counts,' and made the number of accounts that registered 10 or more employees within 14 days of sign-up the only success metric.
- A CS manager phoned 12 customers who had tried the product and then churned. Seven gave the same answer: 'The admin was uploading the list alone in Excel, hit a formatting error, and gave up.'
- The product owner built a KakaoTalk invitation-link flow in two weeks. Over three weeks, they compared it against the existing Excel upload method, alternating 40 new trial accounts per week between the two.
- A pricing-page copy experiment run over the same period reached only 61 samples, so the team left it as inconclusive and recorded it only in the experiment log. The team lead said, 'Even if it stings to log it as a loss, let’s keep it.'
- Six months later, the share of accounts active within 14 days rose from 31% to 52%, paid conversion rose from 9% to 14.6%, and new paid contracts per month went from 12 to 21. Ad spend did not increase at all.