Definition
Growth hacking is the practice of running rapid, low-cost experiments across product, marketing, and user experience to find scalable ways to grow a user base or revenue. Sean Ellis coined the term in 2010, defining a growth hacker as "a person whose true north is growth." The emphasis is on speed of experimentation and measurable outcomes over traditional marketing playbooks.
The canonical examples remain instructive. Dropbox offered 500MB of free storage for every friend a user referred -- a product-embedded growth loop that grew their user base from 100K to 4M in 15 months. Airbnb reverse-engineered Craigslist to cross-post listings, tapping into existing demand. Hotmail added "Get your free email at Hotmail" to every outgoing email. Each was a creative, low-cost experiment that produced outsized results.
Why It Matters for Product Managers
Growth is not just marketing's job anymore. The best growth levers are usually embedded in the product itself -- onboarding flows, sharing mechanics, network effects, and activation optimization. PMs who understand growth experimentation can identify these levers and prioritize them on the roadmap.
At product-led growth companies like Slack, Figma, and Notion, the PM owns the growth model alongside the product experience. Slack grew from 0 to 8M daily active users primarily through product-driven growth: free tier generosity, viral team invitations, and integration hooks that made Slack stickier with every connected tool. A PM who thinks growth is "marketing's problem" would never have prioritized the integrations API that became Slack's most effective acquisition channel.
The shift from growth hacking to growth product management also matters. Early-stage startups benefit from scrappy, ad-hoc experiments. But by Series B, you need a structured experimentation system: hypothesis documentation, statistical significance thresholds, holdout groups, and post-experiment reviews. Duolingo runs 300+ A/B tests per quarter with this level of rigor.
How It Works in Practice
Map the growth model -- Before running experiments, understand your funnel: acquisition, activation, retention, referral, revenue (the AARRR framework). Identify which stage is the current bottleneck. If retention is 20%, fixing acquisition is like pouring water into a leaky bucket.
Generate experiment ideas -- Brainstorm 20-30 experiment ideas targeting the bottleneck stage. Sources: competitor analysis, user session recordings, conversion rate data, support tickets, and analogies from other industries. Quantity matters more than quality at this stage.
Prioritize ruthlessly -- Score experiments on expected impact, confidence in the hypothesis, and cost to run. ICE scoring (Impact, Confidence, Ease) works well here. Pick the top 3-5 experiments per cycle.
Run experiments with speed and rigor -- Ship the smallest possible version of each experiment. Measure against a control group with sufficient sample size for statistical significance. Most experiments fail -- a 20% win rate is considered good. The key is running enough experiments that the winners compound.
Double down on winners, kill losers fast -- When an experiment shows a statistically significant lift, invest in scaling it. When it shows no impact after sufficient sample size, kill it immediately and reallocate to the next experiment. LinkedIn's growth team scaled their "People You May Know" feature after it showed a 30% increase in connections per user in the initial test.
Common Pitfalls
Optimizing vanity metrics -- Growing signups is meaningless if users do not activate. Always connect growth experiments to metrics that correlate with retention and revenue, not just top-of-funnel volume.
Sacrificing user trust for growth -- Dark patterns (forced sharing, misleading notifications, hidden unsubscribe buttons) can juice short-term metrics but destroy long-term trust. LinkedIn learned this the hard way with their aggressive email invitation feature that triggered an FTC complaint.
Running experiments without statistical rigor -- Declaring a winner after 2 days and 100 users is not experimentation -- it is confirmation bias. Define sample size requirements and significance thresholds before launching.
Ignoring second-order effects -- A growth hack that increases signups by 40% but attracts low-quality users who churn immediately is a net negative. Measure downstream impact, not just the immediate metric.
Related Concepts
Product-led growth is the go-to-market strategy that embeds growth levers into the product experience. Activation rate is typically the highest-leverage metric for growth experiments -- improving activation compounds across the entire funnel. The AARRR Pirate Metrics framework provides the funnel model that growth experiments target.
Explore More PM Terms
Browse our complete glossary of 100+ product management terms.