Growth hacking
Also: Growth hacking, Growth marketing, Growth engineering, Data-driven growth, Piratage de croissance, Marketing de croissance
An experimental, data-driven approach to rapid growth by identifying and scaling the most efficient acquisition levers.
What it is
Growth hacking is a disciplined, experiment-led method for driving fast, measurable business growth. Instead of relying on large fixed budgets or intuition, growth teams run many small, cheap tests across the customer journey, keep what works, and scale the winners. The term was coined by Sean Ellis around 2010, but the underlying idea is simply the systematic application of the scientific method to acquisition, activation, retention, and revenue.
It is not a single tactic or a growth "trick." It is an operating discipline that combines marketing, product, data, and engineering into rapid iteration loops.
Why it matters
- Capital efficiency: it optimizes cost per acquired and retained customer, so growth does not depend on ever larger spend.
- Speed of learning: short experiment cycles surface what actually drives growth faster than annual planning.
- Cross-functional alignment: it forces marketing, product, and finance to agree on shared metrics.
- Scalability: validated levers can be scaled with confidence because the underlying economics are already tested.
How it is used in practice
Most teams follow a repeating loop, often called the growth loop or build-measure-learn cycle:
1. Identify a metric to move and form a hypothesis.
2. Prioritize ideas (frameworks like ICE: Impact, Confidence, Ease).
3. Test with a small, controlled experiment (A/B test, landing page, referral offer).
4. Analyze results against a clear success threshold.
5. Scale winners, kill losers, document learnings.
Common levers include referral programs, onboarding optimization, pricing tests, viral loops, SEO content, and activation nudges. The North Star Metric (one number that captures delivered value) keeps experiments focused on durable growth rather than vanity metrics.
Concrete worked example
A B2B SaaS company sees that only 30% of trial signups ever reach their "aha" moment (creating a first project).
- Hypothesis: a guided setup wizard will raise activation.
- Experiment: show the wizard to 50% of new signups for two weeks.
- Result: activation rises from 30% to 42%, and 14-day retention improves 8 points.
- Decision: roll the wizard out to 100%, then test the next bottleneck (trial to paid conversion).
Because the change was cheap and measured, the team scaled a proven lever rather than guessing. The core mindset: treat growth as a portfolio of experiments, not a fixed plan.
See also
Frequently asked questions
What is growth hacking, in one sentence?
Growth hacking is an experiment-led method for driving measurable business growth: run many small, cheap tests across acquisition, activation, retention and revenue, keep what works, scale it. The term was coined by Sean Ellis around 2010, but the substance is the scientific method applied to growth. It is a way of operating, not a single tactic or trick.
What is the difference between growth hacking and traditional marketing?
Traditional marketing usually plans on an annual cycle with fixed budgets and channel commitments; growth hacking works in short test cycles and only scales levers whose economics have been verified. The second difference is scope: growth hacking spans product, data and engineering, not just communication. That is why teams call it growth engineering as often as growth marketing.
Who should learn growth hacking?
Growth hacking is most useful to CMOs, CDOs and general managers who own a growth number and have to defend cost per acquired and retained customer. It also matters to product and finance leaders, since the method forces marketing, product and finance to agree on shared metrics before any experiment is launched.
What are the steps of a growth experiment loop?
Five: identify a metric to move and form a hypothesis, prioritize ideas (with a framework like ICE for Impact, Confidence, Ease), test with a small controlled experiment, analyze against a success threshold set in advance, then scale winners and kill losers while documenting what you learned. The loop repeats on the next bottleneck.
Can you give a concrete example of a growth experiment and its result?
A B2B SaaS company saw only 30% of trial signups reach their "aha" moment, creating a first project. It showed a guided setup wizard to half of new signups for two weeks: activation went from 30% to 42% and 14-day retention gained 8 points. The wizard was rolled out to 100%, and the team moved on to the next bottleneck, trial-to-paid conversion.