Hacking Growth by Sean Ellis & Morgan Brown provides a methodology and playbook for growth hacking, including the team, process, and best practices to make it work.

Growth hacking is the process to acquire and retain users with creative and often low-cost strategies, blending marketing, product development, and engineering skills. Growth hacking takes the lessons of the Lean Startup (which focused on a lean approach to business model and product development) and applies it to customer acquisition, retention, and revenue growth.

Part 1: The Method

Building Growth Teams

The book argues that the traditional organizational structure of companies with product, marketing, design, and engineering being siloed in their own departments results in a lack of communication and collaboration. This lack of communication and focus makes it difficult to come up with creative approaches to drive user and revenue growth. By creating a cross-functional growth team, you're able to more rapidly come up with and execute on creative growth experiments.

Growth Team

These teams should have the following key roles:

Process

The team follows an iterative process where they analyze data, generate ideas to test, prioritize those ideas, and then execute and test the top ones in an experiment. The results of this experiment lead back into analysis where the cycle starts again.

Reporting Structure

There are two common structures for where growth teams report in the organization:

Product-led Model

teams report into a VP of Product
easier to implement when you have established management structure

Independent-led Model

teams report into a VP of Growth
easier to implement in early stage companies

Determining If Your Product Is Must-Have

You shouldn't focus or invest in growth until you have a product that is "must have". An important part of that is the product actually delivering on it's promised value via an "aha moment" -- the moment that the utility of the product really clicks for the users; when users really get the core value (what it's for, why they need it, and what benefit they derive from it). There are multiple ways to determine if you have a "must-have" product.

The "Must-Have" Survey

You can conduct a simple survey with the following question:

How disappointed would you be if this product no longer existed tomorrow?
a) Very disappointed
b) Somewhat disappointed
c) Not disappointed (it really isn't that useful)
d) N/A - I no longer use it.
What would you likely use an alternative to [product] if it were no longer available?
What is the primary benefit that you have received from [product]?
Have you recommended [product] to anyone?
What type of person do you think would benefit most from [product]?
How can we improve [product] to better meet your needs?
Would it be okay if we followed up by email to request a clarification to one or more of your responses?

If you have 40% of respondents reply with "very disappointed", you've achieved sufficient must-have status. If you haven't achieved this, you may need to tweak the product, target a different audience, or put in substantial development.

Measuring Retention

Another important way to measure must-have is via retention rates. These should be compared to your competitive set.

Getting to Must-Have

If you haven't achieved must-have status, you need to understand what your customers need and how you can most efficiently get them to the aha moment. There are three key methods for doing that:

Identifying Your Growth Levers

After ensuring your product is must-have, you then need to scientifically understand how you're going to drive growth -- what your growth levers are and whether they are the right ones to achieve the desired results. To do this, you need to formulate your fundamental growth equation. This is a simple formula that represents all the key factors that will combine to drive your growth. This formula is unique for every product. The way to determine your essential metrics is to identify the actions that correlate most directly to users experiencing the core value of your product. An example for eBay follows:

NUMBER OF SELLERS LISTING ITEMS x NUMBER OF LISTED ITEMS x NUMBER OF BUYERS x NUMBER OF SUCCESSFUL TRANSACTIONS = GROSS MERCHANDISE VOLUME GROWTH

Based on this, you can determine your North Star metric -- the metric that most accurately captures the core value you create. For eBay, it would be "Gross Merchandise Volume". For Airbnb, it would be "Nights Booked". For WhatsApp, it would be "Messages Sent". Having this metric ensures everyone in the organization is focused on what's most important to growth.

Standard analytic tools will give you standard metrics (e.g. sessions, pageviews, events, DAU, etc.). These are distractions. You'll need the ability to gather data that allows you to measure everything in your fundamental growth equation. And equally important is presenting that data in a way that is easily consumable by everyone (or most) people in the organization (e.g. via dashboards).

Testing at High Tempo

Your team should aim to steadily increase the tempo at which they can go through the growth hacking loop. The faster they're able to do so, the more chances they'll have, which leads to more learning and more results. The goal is to learn as quickly as possible.

Preparing

Your first step is to gather team to explain how the process will work, clarify roles for each team member, the methods for generating and prioritizing ideas, what the growth levers and North Star metric are, goals for the tempo of iteration, and an initial baseline review of data.

Stage 1: Analyze

In the first stage, using existing usage data and gathering new data via surveys and interviews, your team will answer the following questions:

What are my best customers' behaviors?

What features do they use?
What screens in the app do they visit?
How often do they open the app?
What items do they buy?
What is their average order size?
What time of day do they shop and on which days?

What are the characteristics of my best customers?

What sources were they acquired from (ad, promotional email, etc.)?
What devices do they use?
What is their demographic background, including age, income, and more?
Where do they live?
How close are they to the store or other stores?
What other apps do they use?

What events cause users to abandon the app?

What screens have the highest exit rates?
Are there bugs that are preventing users from taking a particular action?
How are the products priced relative to other services?
What actions don't they take that users who purchase do?
What is their path through the app, and how much time do they spend in the app before they abandon it?

In subsequent loop cycles, you'll be analyzing the results of experiments.

Stage 2: Ideate

Based on what you learn in your analysis, the team will then generate a broad set of ideas to drive growth. Each idea should be well thought out and include the following:

Idea Name: short name
Idea Description: who's being targeted, what change you'll make for them, where it will be implemented, when will it appear, and how the test will be run)
Hypothesis: simple explanation of cause and effect expected)
Metrics to be Measured: metrics to track to evaluate outcome, including downstream metrics that measure unintended (negative) consequences

Stage 3: Prioritize

All ideas should be scored and prioritized to help the team choose the next set of ideas to test. A simple framework for scoring could be the following (each on a scale of 1 - 10):

Impact: expectation about the degree to which the idea will improve the metric being focused on
Confidence: how strongly the idea generator believes the idea will product the expected impact
Ease: measure of the time and resources needed to run the experiment

Here's an example:

Stage 4: Test

After choosing the experiments for the next testing cycle, the team then moves to implementing those chosen experiments, likely working with other parts of the organization to get them done. If the team discovers that some of the experiments can't be conducted in the next cycle due to dependency or timing issues, they can go back and take another experiment from the queue.

Back to Stage 1: Analysis and Learning

As mentioned earlier, the next cycle's analysis will include results from the experiments that have been run in previous cycles. These can be written up in a test summary that documents what was run, what results the team saw, potential confounding issues that might have skewed the results, and conclusions drawn.

The Growth Meeting

The growth lead will run a weekly meeting (or bi-weekly depending on your tempo) with the growth team that covers the following (basically sprint planning):

Metrics review and update to focus areas
Review of last week's testing activity
Key lessons learned from analyzed experiments
Select growth test for current cycle
Check growth of idea pipeline

Part 2: The Growth Hacking Playbook

Now that you've ensured your product is must-have, organized your data to track your North Star metrics, assembled your team, and have a growth hacking process to follow, you can now execute strategies for the four common areas that drive growth - acquisition, activation, retention, and monetization.

Hacking Acquisition

The main goal here is to find the most cost-effective ways to acquire new customers. Unless you're dealing with a network effect business (social media, marketplaces) where upfront land-grabs at any cost may make sense or where there is a strong competitive presence, in general you have to control your customer acquisition costs. The first phase of work is to ensure two other types of fit: language/market fit and channel/market fit.

Language/Market Fit

Language/market fit is how well the language you use to describe and market your product to potential users resonates with them and motivates them to give it a try. You can no longer dictate what route a user is exposed to your product, so this includes all language on your website, in your web or mobile product, emails, mobile notifications, and print and online ads.

The first message a user sees must convey, in 8 seconds or less, how your product can benefit them to meet a need or desire they have. The language should concisely communicate your product's core value (the aha moment) and answer the simple question "How will this improve my life?".

Language is easy to modify, so it's common to run A/B tests on variations. You generate lots of variations of language (drawing from language your customers use to describe your product, from surveys you've done, from the customer support team, or from interviews), create Bitly links for each, deploy the multiple versions and then measure the results.

Some examples:

"1000 songs in your pocket" - iPod
"Store your phones online" --> "Share your photos online" - Tickle
"Find a Date" --> "Help People Find a Date"
Using trending Google Search terms to direct branding at NastyGal

Channel/Product Fit

Channel/product fit is finding the most cost-effective channel (or channels) for your particular product. You shouldn't, by default, just use the same channels as everyone else (e.g. Google and Facebook paid ads). There's a two step process to finding channel/product fit: discovery and optimization.

Discovery

In this phase, you'll try to find one or two channels that have the best fit. To start, these are the three common categories of channels and example strategies within them:

And then within each of these there are dozens of specific tactical choices. For example, content marketing could include case studies, infographics, e-books, interviews, AMA's, how-to-guides, forums, etc. The options are endless here. So to make a first cut you can use the type of product and business model as a guide. B2B enterprise will generally require a sales team, trade show presence, content marketing, etc. E-commerce requires high-volume, so search ads and SEO are vital. But you don't have to stick to these obvious channels. Here's a guide to help narrow things down:

Once you have an initial list of channels, you then brainstorm and prioritize experiments within those channels based on the following six factors:

Cost - how expensive an experiment would be
Targeting - how easy it is to reach a specific audience with an experiment
Control - how much control you have over the experiment (including after it's live)
Input Time - how much time will it take to launch the experiment
Output Time - how long will it take to get results from the experiment
Scale - how large an audience can you reach with the experiment

You can score these on a scale of 1 - 10 and average the results to help guide your selection of experiments. For example:

Optimization

As was discussed in the process, you'll analyze the results of the experiments and decide which channels to continue to pursue and further optimize. At a certain point, it's possible that you'll reach a natural ceiling with a channel, so it's important to keep trying new things.

Best Practices with Virality

There are several best practices to keep in mind if you're going to pursue virality:

You need a must-have product first.
Word-of-mouth is different than "instrumented virality" where the product itself provides a mechanism for users to hook in more users.
The "instrumented virality" experience of your product should be must-have as well (or at least user-friendly and delightful).
One way to think about viral potential is VIRALITY = PAYLOAD x CONVERSION RATE x FREQUENCY, where payload is number of people each user would reach, conversion rate is how many invitees respond to the invite, and frequency is how often they are exposed to the invite.
With smaller payloads, "double-sided incentives" is often best where both sides benefit from the offer.
Incentives and rewards should be in synergy with your product's core value.
Seamlessly integrate the invite to share (don't just bolt it on).

Hacking Activation

After attracting users to try your product, your goal is to get them to the aha moment as quickly as possible so they can experience value and become activated, i.e. they actively use your product.

Create Funnel Report of Conversions and Drop-Offs to Aha Moment

The first step is to identify each point in your customers' journey toward the aha moment. Map all of these steps out and ensure you can measure each step using analytics. Your next step is to identify which steps in the journey have the most drop-offs. You can do this by creating a funnel report that displays the rates at which people who come to your product are moving on to each key step.

Understanding Why Drop-off is Happening

One method to understand why drop-offs are happening in the funnel is to conduct a survey at the moment of drop-off. If a user's activity indicates confusion (e.g. lingering on a page) or a user has taken the action that other users are not taking, you can pop-up a brief survey to ask them why.

Here are some example questions to ask each group of users (note that you may get the most valuable information from those that actually took the action):

Removing and Adding Friction

Friction is any hindrance preventing someone from accomplishing an action they're trying to complete (e.g. ads popping up, CAPTCHA before completing form, etc.). A way to think about it is:

DESIRE - FRICTION = CONVERSION RATE

By either increasing desire or reducing friction, you can increase your conversion rate. The former is much harder than the latter. There are times when you'll actually want to add friction

Optimize New User Experience

The New User Experience (NUX) is often the best place to reduce friction. There are two rules when designing and optimizing this experience:

Treat it as a unique, onetime encounter (i.e. it's a separate product of its own)
The first landing page of the NUX should accomplish three fundamental things: communicate relevance (does page match intent or desire), show the value of the product (answering question "What's in it for me?"), and provide a clear call to action.

There are a few key tactics that prove useful:

run many experiments with language
simplify the sign-up process using single sign-on
flip the funnel so that your users can experience the joys of your product before asking them to sign-up

Adding Positive Friction

Counterintuitively, it's often helpful to introduce friction into your product. Game designers are experts at this. They keep you in a flow state by presenting actions and challenges that you are capable of completing to help you learn the game and slowly level up to more difficult challenges. By taking simple actions, you make a psychological commitment to make future actions. The more information people put into your product, the more their commitment increases, through a concept called stored value.

Here are a few ways you can apply these ideas:

Learn Flow: Design your NUX to educate new users about the product, its benefits and value, and how to use it.
Upfront Questionnaire: Ask new users about their interests or about the problems they are seeking solutions to immediately creates a form of commitment while also allowing you to potentially personalize the product to their needs.
Gamification: Offering rewards or perks to customers for taking actions that align with the core value they experience.

Triggers

Another way to guide and encourage users to reach the aha moment is via triggers. Triggers are any sort of prompt that provokes a response from people, common ones being email notifications, mobile push notifications, and calls to action on a landing page. To design successful triggers (or prompts), you can use the Fogg Behavior Model. Successful prompts balance how much they motivate users to take an action with how convenient that action is to take. Be judicious in how you use them.

Some common prompts and triggers are as follows:

Account creation: encourages users to complete their account
Purchase messages: encourage purchase with short-term discount
Reactivation campaign: encourage users to come back to re-engage
New feature announcement: share news or updates about the product
Top user incentives: let heavy users know they're special to encourage greater affinity
Activity or status change: change in price, friend taking action, etc.

Here are six principles to craft persuasive triggers:

Reciprocity: people are more likely to do something in return for a favor
Commitment and consistency: people who take one action are more likely to take another
Social proof: people look to actions of others
Authority: people look to people in authority
Liking: people do business with those they like
Scarcity: people take action when worried they will miss out on an opportunity

Hacking Retention

There are significant benefits to managing and increasing user retention, including the following:

extending the duration you have to earn revenue from them
allowing you to invest more in growth
allowing you to learn more about their needs and desires
leads to stronger results from word-of-mouth and viral marketing efforts

There are three phases of retention: initial, medium term, and long-term. But first, you need to be able to identify and track your cohorts.

Measuring and Tracking Retention

You need to determine what metrics you'll use to measure retention. The nature of your product or service usually dictates the frequency with which customers make purchases and what metric you should use (e.g. eCommerce, SaaS, travel, CPG, etc.). Once you've determined your retention metric, you'll break your retention data down more finely using cohort analysis. Cohorts are usually defined by the date they first purchased from you. This could be aggregated by month or quarter or more granularly by week or day. A typical cohort chart looks like the following where it tracks the number or percentage of customers retained for each cohort over time:

Charts like these help you understand what impact your growth efforts are having on retention. For example, if a growth effort resulted in a large increase in acquisition for one cohort but that cohort's retention was low, it might suggest that the campaign attracted customers outside your ideal customer profile (ICP).

Initial Retention

From analyzing your cohort data, you'll see drop-off points in initial retention. You can survey or talk to these customers to understand the causes of defections. The experiments you'll use to address these issues will most likely be very similar to those that you used to improve activation -- getting users to experience value as quickly as possible, using triggers, etc.

Medium-Term Retention

The core goal in the medium term is to make use of your product habitual. Whenever someone wants to buy or use a product of your type, they turn to you rather than a competitor. In other words, they're loyal to you. The key to habit formation is convincing customers of the ongoing rewards they receive from returning to your product. The Hooked Model, often referred to as an engagement loop, is where external triggers serve as prompts to action. Those actions should be rewarded in some way which leads to a further investment in the product. The perceived value of the reward leads to greater retention, so growth teams should experiment with a variety of rewards. The more action that is taken, the greater the reward, and the greater the perceived value.

Rewards should always have incentive/market fit where the reward is tied to the value your specific product provides to customers. Some common rewards are as follows:

savings, coupons, cash vouchers, or gifts
access to special features (e.g. frequent fliers get access to lounges, preferential boarding)
brand ambassador programs that confer social rewards (e.g. Yelp Elite)
recognition of achievements (e.g. sending behavioral emails when customers pass certain milestones, such as congratulatory notification when you record your 10,000th step in a day or completed your longest trip ever or a post receives 100 likes)
customized or personalized experiences (e.g. personalized content)

In addition to rewards, you can also let users know what's on your product roadmap. Users are encouraged to stay with your service in anticipation. But if you fail to deliver on those promises, it could backfire by irritating customers and losing trust.

Long-Term Retention

Over the long-term, there are two strategies to retain customers: 1) optimizing the current set of product features, notifications, and subsequent rewards from repeated use; and 2) introducing a steady stream of new features over a long period of time. In addition, ongoing onboarding is important so that you're moving your users along a long-term learning curve. As you introduce new features, you need to ensure that existing customers are learning the skills that will help them realize value from them. Each new feature or skill should only be encouraged after the previous skill has been mastered.

Users whose activity drops to zero can be added to a resurrection flow where they should be sent a series of email communications designed to win them back, often reminding them of the aha moment or core value of the product.

Hacking Monetization

Your ultimate goal with growth hacking is to increase the number of customers and their lifetime value (LTV), so it's important that growth teams spend just as much time on hacking monetization as they do on acquiring, activating, and retaining customers.

Map & Identify Bottlenecks in Monetization Funnel by Cohort

As with other steps, you should map out the entire customer journey and identify all points along the way where there are opportunities to earn revenue from customers. These points will vary significantly based on the business model (e.g. retail, SaaS, advertising, etc.). But you'll want to measure conversions at each of these points to help identify pinch points or bottlenecks. These pinch points are good starting points for more detailed analysis.

Your analysis should be separated into cohorts based on revenue. For example, high-profit vs. lower-profit customers, subscription plan, spend per year/month/week, engagement (with the product and with ads). You can further subdivide the cohorts by location, age and gender, types of items purchased or features used, acquisition source, device type, Web browser, number of visits per period, date of first action, etc. You should be looking for correlations between these cohorts and revenue. Uncovering these correlations will suggest experiments to run.

This analysis will also help you develop personas of your customer groups.

Ask Customers What Benefits They Want

You should regularly ask your customers what additional benefits they want from your product. You should explore and test these features through surveys and experiments before offering them more widely. Asking customers to stack rank benefits/features can also be useful.

Using Data to Customize Offers

You can use personalization to customize recommendations using complex recommendation engines or something as simple as a Jaccard similarity coefficient that determines how similar two products are to each other. However, you need to be careful to not be intrusive and reveal information about customers that they may not want revealed (e.g. that they're pregnant).

Optimize Pricing

You should be looking to develop persona/pricing fit, where product plans and pricing meet the needs and expectations of the buyers in your audience. You can understand price sensitivity using a survey similar to the Van Westendorp Price Sensitivity Meter (described more thoroughly in Product-Led Growth). You can combine this with the feature research you did earlier (understanding what features customers find valuable).

For each persona you've identified (from your customer cohort analysis), you can understand the features they find valuable vs. less valuable and combine that with their willingness to pay and lifetime value. By comparing these values, you can create feature bundles and price points for plans.

Another method to price is based on your value metric. A good value metric aligns with where your customer perceives value, scales as the customer uses the product more, and is easy to understand.

Pricing Psychology & Best Practices

There are several strategies you can use based on consumer psychology:

Pricing relativity: People's perceptions of prices are influenced by the prices of other options they are offered (e.g. showing magazine subscriptions of "Print subscription" and "Print & web subscription" at the same price point makes people think they're getting a deal).
Price elasticity: It sometimes makes sense to raise prices rather than lower them (you can test this using experiments).
Penny gap: Consumers perceive a large difference between a product being free and having to pay even a small amount for it. It becomes very critical to optimize your strategy for add-ons and upgrades.
Reciprocity: You can try giving something for free just before asking for purchase (e.g. Costco free samples, HubSpot tools, freemium apps).
Commitment & Consistency: Ask users to make small commitments first before making a larger commitment (e.g. adding items to wish list, free trial, etc.).
Social Proof: Effective reviews and testimonials should be credible, relevant, attractive, visual, enumerated, nearby purchase points, and specific (CRAVENS). Add testimonials, logos, results, number of users or shoppers, etc.
Authority: Even the hint of authority is helpful ("I want an expert opinion. Sign me up!").
Liking: Use images of real people in your referral programs.
Scarcity: You can show deals or items that are already sold to show customers what they missed out on.

A Virtuous Growth Cycle

Breakout companies sustain their success and constantly push for more. Some lessons to sustain growth:

Constantly be moving: Encourage teams to constantly experiment and innovate. This will help avoid growth stalls.
Double down: If you have success with a strategy, don't assume that you've maximized the return from it. Continue to invest and innovate on that strategy. Companies often find the most success from doubling down on successful innovations.
Invest in data: You may find that you don't have the data to come up with new hypotheses. It could be worth it to invest in collecting more data to help you uncover new insights.
Experiment with new channels: After focusing on one or two channels and finding success, you can try experimenting with other channels.
Open the ideation process: Involve more teams and members of the organization in ideation.
Take moonshots: Experiment with ideas that could help break your company out of a local maximum.

Hacking Growth