Hacking Growth: How Today's Fastest-Growing Companies Drive Breakout Success

by Sean Ellis

After the success of my growth strategy at LogMeIn, I decided to focus on helping early stage companies accelerate their growth through experimentation. So when Drew Houston reached out to me to discuss how I could work with Dropbox, I couldn’t wait to implement the method that I’d developed. My first step was to get Houston’s buy-in to conduct a simple survey of the current users to calculate what I called (and what you’ll read about in more detail later in the book) the product’s must-have score. The survey asked the simple question “How would you feel if you could no longer use Dropbox?” Users could respond “Very disappointed,” “Somewhat disappointed,” “Not disappointed,” or “N/A no longer using the product” (I wrote the question this way because I found that asking people if they were satisfied with a product didn’t deliver meaningful insights; disappointment was a much better gauge of product loyalty than satisfaction). Having already run this survey at numerous start-ups, I had found that companies where more than 40 percent of respondents said they would be “very disappointed” if they could no longer use the product had very strong growth potential, where those that fell under that 40 percent threshold tended to face a much harder path in growing the business (due to user apathy). When I saw the results of the survey, though, even I was blown away. Dropbox’s score was off the charts, particularly for users who had fully explored the features of the product.

As Facebook’s vice president of growth Alex Schultz puts it: “If you’re pushing code once every two weeks and your competitor is pushing code every week, just after two months that competitor will have done 10 times as many tests as you. That competitor will have learned 10 times, an order of magnitude more about their product [than you].”

Recall that at BitTorrent, the engineers were invaluable in recommending the development of the lucrative battery saver feature.

Growth teams simply don’t work without software engineers being a part of them.

While we should be clear that some growth teams operate without a dedicated member who is a marketing professional, we advocate including a marketing specialist for optimal results. The cross-pollination of expertise between engineering and marketing can be particularly fruitful in generating ideas for hacks to try.

For example, at Inman News, a trade journal for real estate professionals, where Morgan is COO, the growth team includes the email marketing director, because the company’s growth is heavily reliant on email as a channel for acquiring, monetizing, and retaining customers.

A growth team might not include an analyst as a full-time member, but rather have an analyst assigned to it who collaborates with the team but performs other work for the company as well.

Some companies have created multiple growth teams with different areas of focus, such as at LinkedIn and Pinterest, which has four teams dedicated to new user acquisition, viral growth, engagement of users, and activation of newly acquired users.

The process is a continuous cycle comprising four key steps: (1) data analysis and insight gathering; (2) idea generation; (3) experiment prioritization; and (4) running the experiments, and then circles back to the analyze step to review results and decide the next steps. At

So, for example, an in-depth analysis of customer churn (meaning identifying those who recently abandoned the product) might reveal that the people who are defecting haven’t made use of a particular feature of the product that is popular with avid users. That discovery might lead the team to experiment with ways to get more people to try that specific feature out. Or take another example from the work of our growth team at GrowthHackers.com. In looking at our user data, we found that content submitted by the community that included rich media (such as presentation decks from conferences and videos from YouTube) sparked greater engagement and led to more repeat visits from viewers than posts that simply link to stories elsewhere on the Web. So growth team members came up with a series of ideas for adding more rich media items, such as podcasts and videos, to the site.

In terms of the tasks that team members are charged with, team members will still take on specialty tasks according to their area of expertise, sometimes working independently at least at first. For example, the engineers will take charge of any coding needed for an experiment; the designer will craft any design elements needed; the data analyst will work on selecting the sets of users with whom a change will be tested; and the marketing member will take charge of implementing any experiments with promotional channels, such as with a new Facebook ad campaign.

As Mike Schroepfer, Facebook’s CTO, told Fast Company: “I have one hand in the day-to-day and one hand in the future. It’s a little bit crazy-making at times, but it’s important that our core business continues to do well. Because that is what allows us to aggressively invest in these longer-term things.”6

Rascoff has rallied the whole company around the growth mission by establishing a focus for growth efforts that he calls Zillow’s “Play,” a recurring nine-to-twelve-month growth initiative that the entire company aligns around. In 2008, for example, the company realized that it was losing traffic to a rival upstart, Trulia, largely due to Trulia’s smart use of search engine optimization to rank its home listing data higher than Zillow’s in Google search results. So the Zillow executive team decided that SEO would be that year’s Play, and every team in the company was mandated to make becoming world class at search a priority. This involved a major cultural shift, as the company had previously ignored SEO in favor of other tactics. But in the end every team managed to find ways to improve their SEO efforts, and as a result Zillow was able to catch and surpass Trulia, ultimately buying the competitor in 2015 for $3 billion.7

product-led team, for example, might be dedicated entirely to growing the user base for the company’s mobile app, while another might be assigned to helping to drive readers of an online news service to upgrade to a paid subscription. In some cases the mandate of teams dedicated to a particular product will be limited to improving the performance of one aspect of the product, such as activating new users of an online learning software by optimizing the onboarding process, the term for orienting new users on how to use a product. At Pinterest, John Egan runs a growth team that is dedicated entirely to testing the frequency, content, and calls to action in the email messages and mobile push notifications aimed at getting users to come back more frequently. That scope might sound impossibly narrow, but this intense focus allows the team to really drill down on a key component of the company’s growth. For example, in one of their recent growth efforts, the team built Copytune, a highly sophisticated machine learning algorithm, which allowed them to rapidly test dozens of variants of notifications sent to users, across more than 30 languages, and have the software pick winning versions and tee up subsequent tests in search of even better performance. The results of the program have been extraordinary, adding additional people to the site’s monthly active user (MAU, pronounced “mao”) count.

And in fact the Facebook growth team has worked on an incredibly broad range of growth initiatives, from helping to optimize existing products and features, such as improving the sign-up flow for new users, to even building some of its own products, such as Facebook Lite, which was designed to run on a globe with poor data connectivity.

Walmart websites and mobile applications, like the successful Savings Catcher app we described in the introduction. It also leads the acquisition of promising digital start-ups, such as mobile fashion search app Stylr, and social recipe aggregator Yumprint, and works to integrate their technology and talent into Walmart’s digital offerings.

It’s important to emphasize that even when teams are given independent authority, they need the strong backing of top management in order to navigate the internal sensitivities and frictions that can arise between product, marketing, design, and engineering specialists who have their own notions of what’s important and the “right way” to do things.

At BitTorrent, marketing’s original intent was to concentrate solely on user acquisition. Data analysis was done by the company’s data team, specifically at the request of the product teams, and experimentation had no home and had mostly fallen by the wayside. So when the growth team came in and knocked down these divisional silos, it took some adjustment.

A final cause of potential strife lies in the fact that when you bring together people from such a diverse range of fields and backgrounds, there are almost certain to be differing and sometimes conflicting perspectives and priorities. Engineers tend to be most interested in working on the most technically challenging jobs, whether or not the solutions they come up with will have a meaningful impact on growth. Product managers are typically obsessed with product development and launches and can be infuriated by marketing and sales teams making last-minute requests for changes with poor business rationales. User experience designers often resist the introduction of experimental features for testing because they don’t want to upset happy users with annoyances. Marketers can become focused on vanity metrics such as the number of website visitors or leads and lose sight of the need to drive up the performance indicators in other parts of the funnel

Another way to mitigate conflict is to ensure that the decisions about which growth hacks to prioritize (and the evaluation of how successful they are) are made strictly on the basis of hard data rather than assumptions, or what Chamath Palihapitiya, the original leader of the Facebook growth team,

At Qualaroo, for example, growth was held back for a time by company lore that maintained we couldn’t raise our prices and still succeed. Yet when we conducted pricing tests, the data showed that we could raise prices by more than 400 percent and still grow by attracting a new type of clientele.

And when experimentation is data driven, team members also generally respect the rigor of the learning process and appreciate the latitude it gives them to rack up a string of failures in order to achieve a win.

As companies grow and evolve, so should their growth teams. At Facebook, the growth team has exploded from its early days of five people to a now sprawling unit with multiple focus points such as international and emerging markets phone units.

GROWTH HACK TO START GROWTH HACKING Implementing the growth hacking process can seem daunting. Creating a cross-functional team can be tricky, as managers of groups may push back about rededicating the time of some of their staff. The notion of so much experimenting can also be uncomfortable for people. Inevitably, there will be naysayers and resisters. The good news is that there is also a virtuous growth cycle in the adoption of growth hacking. A small team with a narrow focus that begins running the growth hacking process and generates a series of wins can spark growing enthusiasm for the process around a company. Once people see the power of the data-driven approach to experimentation—and the growth ideas that come out of it—enthusiasm for the process tends to be infectious.

Implementing growth hacking across a company or even a department won’t happen overnight; so think about starting with a team working on just one product, and perhaps even just one important aspect of how it is being adopted, such as the sign-up page on your website. Or you could create a team to work on optimizing the company’s customer acquisition in a single channel, such as Facebook, or improving the readership of the company’s blog, or the performance of the company’s email marketing.

ALL FAST-GROWTH companies share one thing in common. Regardless of who their customers are, their business model, and the type of product, industry, or region of the globe they’re operating in, they all make a product that a large group of people love. They’ve built products that, in the eyes of their customers, are simply must-have.

Just think of Google Glass and Amazon’s Fire Phone—both innovative products … that nobody wanted.

Microsoft Zune media player, launched in November 2006, which Microsoft reportedly spent at least $26 million to promote but which never generated more than a tepid response.1 The Zune was not a bad product; many critics considered it quite well designed. But it added no “wow factor” to make it more appealing than Apple’s already ubiquitous iPods. Despite continued efforts to stoke sales, including the release of an improved version, the Zune HD, in 2009, the Zune was never able to garner more than a single-digit share of the market and was discontinued in 2011.

One of the cardinal rules of growth hacking is that you must not move into the high-tempo growth experimentation push until you know your product is must-have, why it’s must-have, and to whom it is a must-have: in other words, what is its core value, to which customers, and why.

But the hard truth is that no amount of marketing and advertising—no matter how clever—can make people love a substandard product.

Remember that viral word of mouth can work two ways; it can supercharge growth or it can stop it in its tracks.

A pernicious misconception about growth hacking is that it is primarily about building virality into products.

Sometimes it’s a feature or user experience that was built into the product that is quite different from what was hypothesized in the original product vision as the core value; other times it’s one that was built into the product somewhere along the way as almost an afterthought. Whichever the case, it’s up to the growth team to find out. In this chapter we’ll learn how.

To hasten the app’s viral growth, the team, led by Zack Onisko, devised a brilliant hack of Facebook’s invite system to get more users to share the app with their Facebook friends. At the time, Facebook allowed users who installed a new app to invite other friends to use it, too (and, in fact, many of the apps that grew virally on Facebook took advantage of this mechanism, such as the wildly popular game Farmville).

Facebook only allowed you to invite a maximum of 50 friends at once, and the BranchOut team knew that the conversion rate for Facebook invites was extremely low. The only sure way to drive viral growth, they decided, was simply to prompt users to send more invites. Onisko says they tried hundreds of experiments for how to do so, until they “stumbled upon” a solution. The team had figured out how to allow users to overcome the 50-invite limitation by repeatedly clicking the Next button in a specially designed window, which triggered Facebook’s invite system to suggest another 50 of the user’s friends, and then yet another 50. The tactic supercharged referrals, and BranchOut grew from four million users to twenty-five million in about three months.

BranchOut’s founder, Rick Marini, conceded in a 2012 talk that the company had erred in trying to rush user acquisition without delivering on the product experience.

“What we’ve learned is that there are times when you can get some spike of virality, but if you really want that long-term major user growth it’s got to start with a good product.

Despite raising nearly $50 million in venture capital, BranchOut never became more than a viral one-hit wonder; its wild ride ended in a resounding thud when the company sold its assets to a relatively obscure human resources company, 1-Page, for $2 million in cash and some stock.8

The team ran some tests to see how visitors would respond when reviews were put front and center, and when they saw good results, they pivoted from the original business model of friends asking friends for business recommendations and put the reviews at the heart of the experience. But they didn’t stop there. The team then created 20 million profiles for small businesses in the Bay Area on the site and encouraged users to add their reviews. Growth took off. Meanwhile Citysearch is now little more than a footnote, having been folded into CityGrid Media in 2010.

For Yelp, that experience was the ability to discover promising local restaurants and businesses through trusted community reviews.

For eBay, the aha moment was finding and winning one-of-a-kind items at auction from people all over the world. For Facebook, it was instantly seeing photos and updates from friends and family and sharing what you were up to. For Dropbox, it was the concept of easy file sharing and unlimited file storage. Or take Uber’s aha moment, which Uber cofounder and CEO Travis Kalanick explained as, “You push a button and a black car comes up.

Twitter struggled to sustain growth in its early days until it learned (from doing extensive analysis of its user data) that users who quickly started following at least 30 other users were much more engaged and likely to continue using the service. Digging into why following 30 people seemed to be the tipping point, the Twitter growth team found that getting a steady stream of news and updates from people they were interested in was the aha moment for people. Following 30 people created a stream of updates that made the service “must-have.”

Similarly, at Qualaroo, the website survey company where Sean and Morgan worked together, we identified that trial users who received 50 or more responses to one individual survey were three times as likely to pay for their subscription at the end of their trial when compared with users who did not get a survey with 50 responses during the same period. Here, 50 responses was what it took to deliver the aha moment of seeing how the product delivered new and valuable feedback. At Slack, a team chat and messaging product designed to eliminate internal corporate email threads (and one of the fastest-growing business applications of all time), data showed that once team members had sent and received 2,000 messages to one another, the team became far more likely to make Slack a core part of their communication workflow and upgrade to a paid plan with premium features.

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 Interpreting the results is simple enough; if 40 percent or more of responses are “very disappointed,” then the product has achieved sufficient must-have status, which means the green light to move full speed ahead gunning for growth. Many products won’t hit the 40 percent threshold, however, and in that event, the growth team’s first efforts must be focused on determining why the product isn’t getting a better response. If 25 to 40 percent of respondents answer “very disappointed,” then often what’s needed are tweaks either to the product or to the language used to describe the product and how to use it. If less than 25 percent answer “very disappointed,” it’s likely that either the audience you’ve attracted is the wrong fit for your product, or the product itself needs more substantial development before it’s ready for a growth push.

The question about which type of person users think can benefit the most from the product can help the team focus on a better-defined customer niche and thus target those potential users more effectively. For example, at Inman, where Morgan works, they asked the question of users of a training product for real estate professionals that the company had recently launched, and the answers pointed to it being particularly well suited for new agents. Morgan’s team used this feedback to improve the marketing and advertising targeting to focus on this group of potential customers.

Clearly, the larger the user base when you conduct the survey, the more reliable and informative the information will be. You’re looking to get at least a few hundred responses to the first question to be a reliable guide for this kind of survey.

Somewhat ironically, it’s best if you target the survey at active users rather than those who have gone dormant.

For one thing, once your growth has taken off, it’s not a good idea to even suggest to your customer base that the product might be discontinued by asking them how they’d feel if it were no longer available.