Sean Ellis coined the term "growth hacking" in 2010 after running the early-growth functions at Dropbox (where he was the first marketer in 2008), Eventbrite, and LogMeIn (which IPO'd partly on the strength of his growth work). Morgan Brown ran growth at Inman News, Qualaroo, and Facebook. "Hacking Growth" is their codification of the operating system for cross-functional growth teams: how to staff them, how to find the activation moments that drive retention, how to design experiments, how to instrument the funnel, how to run the meeting cadence that produces compounding results. The book is structured as a working manual rather than a thesis. The central argument is that growth is a discipline, not a tactic, and that companies that build the discipline (with the team structure, the experiment cadence, the instrumentation, and the executive backing) compound far faster than companies that run growth as a marketing campaign. For a CMO inheriting a growth team or building one from scratch, this is the canonical operating reference. The case studies (Pinterest, Airbnb, LinkedIn, Hotmail, Dropbox) are now well-known; the value is in the operating system that produced them.
Core frameworks
1. The growth team structure
A cross-functional team with marketing, product, engineering, data, and design reporting into a single growth lead. The team owns specific metrics (activation, retention, referral, revenue), runs experiments against them, and has decision authority over the product surfaces they touch. The traditional "marketing department asks engineering for changes" structure is too slow to compound because each requested change goes through engineering's separate prioritization process.
Facebook's early growth team under Chamath Palihapitiya is the book's worked example. The team was structured with product managers, engineers, data analysts, and designers reporting into a growth lead, not into the broader product or marketing functions. The team's authority extended to shipping product changes (not just marketing campaigns), which is what made specific metrics movable. The team focused on "seven friends in ten days" as the leading retention indicator, then engineered the product flow to drive new users to that threshold faster. The team shipped changes to the friend-finder import, the new-user signup flow, the people-you-may-know recommendations, and the notification triggers. Each change was a small experiment; the cumulative effect was the user-growth curve that took Facebook from college users to global scale.
The structural contrast: a traditional marketing team that wants to improve activation has to file engineering tickets, wait for prioritization, and accept a multi-quarter lag between idea and ship. A growth team owns the activation surfaces directly and ships in days. The compounding rate difference is substantial.
How to operate: if your growth function is structured as marketing asking engineering for help, you do not have a growth function. Restructure to give the growth lead direct ownership of activation, retention, and referral surfaces. Hire or borrow the engineers and designers into the growth team rather than treating them as a shared resource.
If you only remember one thing: structure determines speed. Cross-functional growth teams compound faster than the marketing-asks-engineering pattern.
2. The North Star Metric
Every growth team needs a single metric that captures the customer value the product delivers. Pageviews are not a North Star (you can goose pageviews without improving the product). Time-spent-on-site might be, depending on the product. For Airbnb the North Star is nights booked. For Pinterest it is weekly active pinners. For Slack it is messages sent per active workspace. The North Star aligns the team and protects it from chasing vanity metrics.
Brown describes Facebook's North Star evolution. The team initially optimized for monthly active users (MAU), which produced growth but masked engagement quality. The team then shifted to daily active users (DAU), which surfaced engagement quality but still allowed engagement-quality issues to hide. The team finally settled on specific engagement metrics (time spent, posts shared, messages sent) that better predicted long-term retention. The North Star evolution mattered because each shift changed which experiments the team ran. Optimizing for MAU produces sign-up funnel work; optimizing for DAU produces retention work; optimizing for engagement quality produces feature work. The metric drives the work.
The qualifying properties of a good North Star: it captures the customer value (not just business value), it predicts long-term retention, it is observable in product data, and it cannot be gamed without actually improving the product. A metric that fails any of these tests is a vanity metric and will produce growth that does not stick.
How to operate: define your North Star Metric and align the whole growth function around it. If you cannot name it in one sentence and defend it against the four qualifying properties, the team is chasing vanity.
If you only remember one thing: the metric drives the work. Choose the metric deliberately.
3. The activation aha moment
Every successful product has an activation moment: the point where the user experiences enough value to become a retained customer. Identifying the aha moment and engineering the new-user experience around it is the single best growth lever.
The canonical examples are Chen-like in their specificity. Facebook's "seven friends in ten days" was the aha moment that predicted retention. Users who reached seven friends within their first ten days were substantially more likely to remain active long-term; users who did not reach the threshold churned at a high rate. Dropbox's was uploading the first file. Twitter's was following thirty accounts. Slack's was sending two thousand messages within a single team workspace. Airbnb's was completing the first booking. Each of these was identified by analyzing the behavior patterns of retained users versus churned users in the first week.
The analytical method: cohort retention analysis. Split your user base into retained and churned cohorts based on a specific retention window (often 30 days). For each cohort, examine the early-stage behaviors (first session, first week). Identify behaviors that strongly correlate with retention. The behavior with the strongest correlation is your candidate aha moment. Test by engineering an intervention that drives new users to that behavior faster; if retention improves, the aha moment is real.
Once the aha moment is identified, the entire activation flow can be optimized to drive users to it faster. The new-user onboarding sequence becomes a structured push toward the aha moment, not a tour of features. The friction-removal work becomes specific: what is keeping users from reaching the aha moment in their first session? Each friction point removed produces measurable retention lift.
How to operate: run the cohort retention analysis. Identify the behavior that distinguishes retained users from churned users in the first week. Engineer the new-user experience to drive that behavior. Measure the lift.
If you only remember one thing: every retained product has an aha moment. Find it. Engineer for it.
4. The high-tempo experiment cadence
Growth teams run dozens of experiments per week, not one campaign per month. The cadence is enabled by lightweight experiment design, fast instrumentation, and a weekly meeting structure that reviews results and queues the next batch. The book prescribes a specific four-stage cycle: analyze (review last week's results), ideate (brainstorm new experiment candidates), prioritize (score candidates by impact and effort), test (ship the prioritized experiments).
Booking.com is the book's most-cited cadence example. The travel platform famously ran over 1,000 experiments simultaneously at peak, with hundreds of small experiments running on different pages and user cohorts at any moment. The cadence compounded into the dominant travel platform because each small win stacked. A 1 percent conversion lift on a checkout page, multiplied across millions of bookings, produced revenue impact that justified the entire experiment infrastructure. The team had built the instrumentation to run experiments cheaply, which made the cadence sustainable.
Most teams cannot match Booking's experiment volume but can match the structure. A weekly meeting with a fixed agenda (review last week's experiments, ideate this week's, prioritize, ship) produces compounding results even at modest experiment volumes. The discipline is the multiplier; the specific volume is secondary.
ICE scoring (Impact, Confidence, Ease) is the book's prioritization shorthand. Each candidate experiment gets a score on three dimensions: how much impact would this have if it works (1 to 10), how confident are we it will work (1 to 10), and how easy is it to ship (1 to 10). The product of the three scores ranks the experiments. The discipline forces the team to skip the low-confidence, low-impact experiments that often consume disproportionate effort.
How to operate: build a weekly experiment meeting with a fixed agenda. Score candidates with ICE or a similar lightweight method. Ship the top-ranked experiments. Review results next week. The cadence is the multiplier.
If you only remember one thing: experiment volume compounds. The team that runs ten experiments per week beats the team that runs one campaign per month.
5. The full-funnel optimization frame
Growth is not just acquisition. The book frames the funnel as acquisition, activation, retention, referral, revenue (Dave McClure's AARRR pirate metrics, which the authors adopt and extend). Each stage has different levers and different experiment types. Most teams overspend on acquisition; the higher leverage is usually in activation and retention because retained users compound over time.
Brown describes a SaaS case where the team's acquisition budget was three times the activation team's budget, even though activation improvements were producing four times the ROI per dollar invested. Rebalancing toward activation produced faster growth at the same total spend. The pattern is common: marketing teams have historical comfort with acquisition spend (paid media is the default lever), while activation work (onboarding, product-led growth, friction removal) requires different skills and different infrastructure. The historical comfort produces over-investment in the wrong stage.
The math of the rebalance: an acquisition improvement that adds 10 percent more users at the top of the funnel produces 10 percent more revenue, assuming everything downstream stays the same. An activation improvement that doubles the percentage of new users who reach the aha moment can produce a 50 percent or larger improvement in retained user count, which compounds month over month. The activation lever often pays back the rebalance within a single quarter.
How to operate: audit your funnel spend across acquisition, activation, retention, referral, revenue. Most teams over-fund acquisition. Reallocation usually produces faster growth than budget increases.
If you only remember one thing: activation and retention compound. Acquisition is linear. Spend where it compounds.
Actionable takeaways
- Define your North Star Metric and align the whole growth function around it. If you cannot name it in one sentence, the team is chasing vanity.
- Identify your activation aha moment using cohort retention analysis. Find the behavior that distinguishes retained users from churned users in the first week. Engineer the new-user experience to drive that behavior.
- Build a weekly experiment meeting with a fixed agenda: review last week's experiments, ideate this week's, prioritize, ship. The cadence is the multiplier.
- Instrument the full funnel before running experiments. Without instrumentation, experiment results are guesses. Spend the engineering time upfront.
- Audit your funnel spend across acquisition, activation, retention, referral, revenue. Most teams over-fund acquisition. Reallocation usually produces faster growth than budget increases.
What this book is NOT about
This book is not a strategy text. It does not tell you what business to be in, what product to build, or how to position. It assumes you have a product that delivers value and shows you how to find more users for it faster. If your product does not retain, no amount of growth hacking will save you; the activation aha moment will not exist to engineer toward.
Two specific misreads to avoid. First, "growth hacking" is not "marketing tricks." The pop-press version of growth hacking that emphasizes viral mechanics and clever campaign hooks misses the operating-system substance of the book. Real growth hacking is the team structure, the experiment cadence, the metric discipline, and the funnel instrumentation. The tricks are downstream of the system. Second, the case studies are not the framework. Many readers extract specific Dropbox or Hotmail tactics and try to apply them directly to unrelated businesses. The tactics rarely transfer; the operating system always does.
Field updates since publication: the book is a 2017 document. Some of the channel-specific tactics (Facebook ads targeting, viral mechanics that relied on platform open graphs, App Store optimization) have shifted. The operating system (team structure, experiment cadence, full-funnel frame, AARRR) is timeless. The case studies are now well-known to the point of cliche; that is the cost of being a foundational book. The most relevant contemporary critique: the rise of product-led growth has shifted some of the activation work from marketing-driven to product-driven. The book anticipates this shift; the structural frame still applies. Pair with Andrew Chen's "The Cold Start Problem" (2021) for the network-effects layer and with Eric Ries' "The Lean Startup" (2011) for the product-market-fit foundation.
Want more?
Borrow the full book on archive.org: https://archive.org/details/hackinggrowthhow0000elli
The original is about 320 pages with extensive case studies and templates. The summary above captures the operating-system frameworks. Read the full book if you want the worked examples, the activation-moment analysis methods, or the team-staffing templates. Pair with Andrew Chen's "The Cold Start Problem" (2021) for the network-effects extension and with Eric Ries' "The Lean Startup" (2011) for the product-market-fit foundation.