Growth hacking is failing most of us — because most of us don’t understand that what works for companies with a growing customer base doesn’t work for earlier-stage companies still looking for product-market fit; and vice versa.

There are actually two very different types of growth hacking, and if you’re not using the right method, you could well sink your company.

Take Andy Johns and Sean Ellis, for example. They’re two of the most highly regarded growth hackers in the world, and yet they’re completely different people who work in completely different ways. Their different approaches reflect why it’s so hard to define what a growth hacker is.

Andy Johns, formerly of Twitter and Facebook and now the highly regarded head of growth of Wealthfront, was raised as a farmer in the Central Valley of California. He speaks with the cadence of someone experienced in patiently husbanding cash crops to market. He’s preternaturally calm, disciplined, a numbers-oriented data scientist.

AI Weekly

The must-read newsletter for AI and Big Data industry written by Khari Johnson, Kyle Wiggers, and Seth Colaner.

Included with VentureBeat Insider and VentureBeat VIP memberships.

Sean Ellis, founder and CEO of marketing technology company Qualaroo and the popular GrowthHackers.com community, was the first marketer at Dropbox, Lookout, and LogMeIn and is widely credited with coining the term Growth Hacker five years ago. He is Johns’ cognitive opposite. Ellis is a fast talking polymath who started his career in direct sales before there was an Internet to advertise on. His tenacious “hustle” to execute, his leaps of insight in leveraging new channels and technology, and his shark-like constant motion is textbook entrepreneur material — an enormous asset to any early stage company. Indeed, he was part of the team that started and sold one of the earliest big players in the ads market — Uproar — to Vivendi Universal after getting it to a peak valuation of $1 billion.

Unsurprisingly Ellis is a serial entrepreneur, and Johns is now leading growth inside yet another unicorn.

growth gridWhere Johns makes small changes on enormous sites like Twitter to get 30,000 more visitors a day, Ellis is constantly on the lookout for the next arbitrage, the next way to stay ahead of his competition, the next channel to leverage, the next idea to tweak into a creative asset.

In gold rush terms, Ellis is the prospector, and Johns is the miner. And it turns out the two archetypes are a perfect match for the two most important company stages: startup and growth. Prospector growth hackers do best in early-stage companies that are still working to find perfect product-market fit and that don’t yet have their first big money invested. Miner growth hackers do best in post product-market fit conditions, when their skills are needed to excavate datasets in large and/or fast-growing businesses and find ways to turn those insights into even more growth.

Put the wrong growth hacker in the wrong company, and disaster happens. Send a Miner to prospect, and she might spend years digging patiently for decades in the Arizona desert, never striking gold. Conversely, if you send a Prospector to mine the gold, she’d probably spend half the resources looking for silver instead.

Still, despite their different approaches, Prospectors and Miners have seven key characteristics in common:

1. Data-driven: Growth hackers make all decisions based on data. Prospectors evaluate their results with data, but Miners also use data to decide which problems to solve. By definition a Prospector may not know where the gold is, so she has to take chances and try things, sometimes with very little data, to see if there’s a vein of gold. But she uses data to evaluate the success of that test. Work decisions thus become not about WHO is right, but WHAT is right. This makes for a faster and less political decision-making process, a critical factor for increasing speed in companies of all sizes.

2. Technical aptitude: In tech companies, both Prospectors and Miners are friends of engineers because the website is the company and nothing gets done without developers. Good growth hackers have a strong aptitude for working with engineering via agile methods, design skills, and technical knowledge. Neither Ellis nor Johns is an engineer. But like Steve Jobs, they learned enough about engineering to understand how to work with developers effectively. In a non-tech company, the technical aptitude needs to be in the areas required to effectively implement the tactics required.

3. Customer empathy: We’re talking about marketing, after all. So every growth hacker cares about customers. Prospectors talk to every single one of the first 1-100 or so customers to understand where the product must evolve to meet the customer expectation. Miners, well, they mine customer and usage data to create better experiences. “Developers are so busy coding,” Ellis notes, “that they often have little perspective on the real-world experiences of the customers using the product.” A great growth hacker’s job is always bridging that gap.

4. Team orientation: Because decisions are made with data and don’t rely on intuitive leaps of individuals, the best growth hackers don’t have big egos. No growth hacker has the time or energy to come up with all the ideas and implement them. They need to be able to work within teams to design, execute, measure, and improve key performance indicators. Because the data measures success, growth hackers have a mindset of “how can I help,” not “that isn’t my job.”

5. Curiosity: Good growth hackers must have that strong sense of “what if” that is common in science. Prospectors will be curious about early customers’ similarities and about trying the most recent and creative tactics, riffing off case studies and testing new platforms. Miners will be curious about data and what it reveals in terms of opportunities. Whether it’s teaching themselves to code, deep-diving into data analysis, learning photoshop, or a tackling a new marketing platform, growth hackers at all levels should be eager to tackle any challenge the business presents.

6. Focus/Discipline: Both Prospectors and Miners need to be focused and disciplined. Time is a factor in both roles. Prospectors can’t check every square inch of earth, and Miners can’t chase every vein of gold at the same time. This also means it’s critical for growth hackers to have solid project management, follow through, and detail orientation skills — or be supported by team members who do.

7. Grit: Just like entrepreneurs, growth hackers have to have grit. A hypothesis-driven approach can mean many, many points of failure. In the early days, success may seem very far away indeed. Think about AirBnB founders sleeping on air mattresses and eating cereal for years. Miners have to have the grit that comes from knowing not every change will be a hit and that big changes will take time to socialize.

Up next:

Pg. 2: How to be a Prospector-style growth hacker

Pg. 3: How to build out a Miner-style growth hacking operation

prospecting for growth

How to be a Prospector-style growth hacker

From Wikipedia: Prospecting is physical labour, involving traversing (traditionally on foot or on horseback), panning, sifting and outcrop investigation, looking for signs of mineralisation.

Here’s how that translates into a Prospector role at an early-stage startup:

1. Forget the playbook

Sean Ellis started out as the ad sales person for a very early company that would go on to become Uproar. When he realized there weren’t enough customers to sell to, he pitched the idea of working on finding those customers.

Having no traditional background in marketing, Ellis had no preconceived ideas of how marketing should be done. So he brought his funnel-oriented mindset from sales directly into marketing. Ellis calls it “the purity of cluelessness.” Reid Hoffman, the founder of LinkedIn, likes to say that every startup marketing plan is sui generis — that a playbook doesn’t exist.

Are you an experienced marketer? Get ready to be humbled. According to Ellis: “At least half of your assumptions are wrong.”

So you can’t follow Ellis’ playbook, or even your last company’s playbook. You have to make your own, and it will differ at every company you work for.

2. Take a systematic approach to growth

“You just have to be systematic and disciplined about [growth efforts] in order to explore the best possibilities,” says Ellis.

That doesn’t mean you should go make a traditional marketing plan. “A marketing plan for a startup is silly,” he says. He suggests the outline of marketing efforts should be called a “marketing hypothesis” in the early days, to signal the extent of what is unknown.

But if you don’t put a stake in the ground, you can find yourself drifting with every shiny new idea that comes in. Ellis suggests quarterly themes with monthly goals.

3. But be ready to abandon that approach if things change

Companies shouldn’t commit too early to a heavy-duty project management framework. As someone who knows the downsides of chasing new and shiny things, Ellis has launched a new project, Canvas, that will help startups work through the cycle of planning in a more efficient way.

“Part of what I’m trying to do with Canvas is answer the question: “What should replace the marketing plan? The idea of a fixed marketing plan even on a quarterly basis doesn’t make sense when channels get saturated and change so quickly. You need a living system to enable collaboration and experimentation so you are in constant learning mode and evolving, and you are replacing channels faster than channels are fading.

4. Focus on acquisition, not awareness

“Awareness building is a waste of startup resources,” says Ellis. By awareness, he means paying for ads or investing time in content marketing with the sole purpose of letting people know you exist, not gaining customers. These are the parts of the profession marketers call “promotion,” or that developers call “shouting.” This is only a small slice of what marketing does, and at this stage, Ellis believes it’s a waste of time and money, because a) awareness gives you no feedback loop to iterate on, and b) it’s unrealistic for a startup to track ROI of awareness building given its limited resources.

So, instead of the traditional AIDA framework he learned in sales (explained in this classic Glengarry Glen Ross scene with a deliciously foul mouthed Alec Baldwin); where the first part of customer contact is awareness, Ellis prefers 500 Startups founder Dave McClure’s “AARRR Startup Metrics for Pirates” framework. 500 Startups is uniquely focused on helping startups to grow; it has an in-house marketing team to support its portfolio companies and a “Distro Fund” designed specifically to fuel growth once product market fit is found. Both Ellis and Johns are mentors to 500 Startups portfolio companies.

In the AARRR framework, awareness (which is expensive to buy in today’s world, as discussed in part one of this piece), is replaced by acquisition and activation — a combination of getting people to the site and getting them to sign up, or convert, so they can proceed to the next phases: retention, referrals, and revenue.

In my post last month about why so many companies are failing at growth hacking, we met Traba CEO Terrence Cummings, who had been told by would-be investors to spend money on ads. He took that advice and invested nearly $10,000 in awareness for his company to lower the investors’ perceived risk of the company. This is what awareness testing costs today in ads alone, not counting any opportunity cost of time and the cost of hiring specialists either as contractors, agencies, or employees. “To get to a point where you can use your [awareness] budget wisely, you need to spend a significant amount of money, which early stage startups do not have,” says Cummings. By the end of summer 2014, the awareness efforts were having sub optimal results, both founders were poorer, and the company still hadn’t raised its next round. Worse, according to Cummings, its growth and engagement actually dropped.

“We learned that [in the early stage] we couldn’t rely on anyone but ourselves,” Cummings says. He says if the company had continued to take would-be investors’ advice to tackle awareness, it would have gone out of business. Ultimately, it turned off all paid advertising, regrouped on its mission and purpose, and focused on Twitter advertising to draw the right top-of-funnel customers in, which cost zero hard dollars.

Traba’s growth is now back to five times what it was when it was using paid ads. Its advice to bootstrappers? If you decide to pursue awareness tactics, be disciplined. “There are a lot of great paid [awareness] products out there to grow your company, but each one takes a significant amount of time and is a distraction from the focus needed to build a great product.”

Cynthia Schames, founder and CEO of Abbey Post, had a similar experience. “One of our investors pushed hard for us to go into Adwords.” But, she says, “Search engine marketing does not work for a considered purchase. It works for commodities, and so we pulled the plug on paid ads and turned to Pinterest, which wasn’t charging us.”

Ellis’ approach is a refreshing alternative: “I talk to the prospects on our pre-product-market-fit product  every single day. One of the main reasons I created a waiting list for the product was so that our VP of Product or I could personally demo the product to every single person who signed up, get their feedback, and uncover their reasons for trying it.”

Ellis even rejects startup wisdom to avoid big companies as customers at first. Reflective of his focus, he has very clear goals for the time he spends with them. “I have no expectation [Canvas] is going to be good enough [for a big company right now]. The benefit of talking to them now is to find out what motivated them to sign up for my waiting list.” Ellis then uses that information to refine both his marketing pitch and his product to meet those expectations over time. Note that he’s not selling. If he tried to do that, he’d become distracted and pull the company off course.

5. Know that every successful channel is a dying channel

The lifecycle of marketing channels is accelerating. After you find a vein of gold, you could finish mining it in weeks or months, so you need to constantly be on the hunt for the next opportunity. According to Ellis, the goal of all marketers should be to replace every successful channel before it dies. GrowthHackers.com’s backlog of product and marketing tests now has over 500 items on it.

Inspiration can come from anywhere. In the early days of Uproar, Ellis noticed how successful the affiliate channel was for Amazon. He took that idea, tweaked it for his unique situation, and created the idea for a game, but as an embeddable unit like an affiliate store instead of in a banner ad with higher CPM (cost per thousand impressions). The result was a cost per acquisition of $0.50, a small fraction of what other companies were paying for a free registered user. Many of those affiliates were tiny sites, which underscores the point that there are no silver bullets in startup marketing.

GrowthHackers.com has hired a product manager whose role is to take those 500 ideas mentioned above and run a weekly growth meeting, where the core team members, including Ellis, picks 3-5 ideas to be implemented and tested that week.

What’s a good upper boundary for test volume? Hubspot’s Sidekick team has publicly shared numbers that suggest they’re running 30 or more tests per week.

growth hacker characteristics6. Don’t look for unicorns — growth is done in teams

“The profile of the people who are good at [growth] is very small,” says Ellis. He’s given up on hiring what he calls unicorns. Hiring for this skill set is such a bottleneck that Ellis created the term “growth hacker” to address it. “I kept getting resumes from brand marketers who just didn’t get the data-focused, scientific method we were using,” he said. “I was forced to come up with a new definition so I could attract a different kind of person.”

Ellis has a point. An unscientific Google search yielded 17,500 results for people with the term “growth hacker” in their profiles on LinkedIn. You can see the list here. Given that the title is self imposed, it’s fair to assume the pool of experienced, qualified growth hackers is even smaller. With over 118,000 jobs for startup and growth marketing professionals on LinkedIn, that means there are at least six jobs out there for every self-identified growth marketer.

Ellis now solves this problem by building growth into his entire team from the beginning.

“When you hire an individual, they have to be everything,” Ellis says, recommending the CEO complement herself with others who have different cognitive abilities. For example, Ellis says he’s an ideas guy who sometimes struggles with execution, so his growth product manager is obsessively execution focused.

Up next:

Pg. 3: How to build out a Miner-style growth hacking operation

Mining

How to build out a Miner-style growth hacking operation

In gold-rush days, once gold was found in “them thar hills,” miners would rush in to extract it from the vein as soon as possible, while fending off all kinds of claims and attacks on their territory. Large scale mining required fast and innovative technology to stay ahead of the rush. And it took significant financing to get the most out of these advanced mining techniques, just like you do today in high growth technology startups. That demand for financing signals the need for a change in most companies’ growth cultures.

Here’s how you make that change:

1. Let your growth DNA evolve

In the early stages of successful companies, according to Andy Johns, growth DNA is often driven by idealism and a strong product philosophy, usually led by a developer or product person.

He says that mentality has to change once the company hits growth stage. “At scale you have to accept that decisions can be made in other ways, and that’s where [miner-style growth hacking] becomes a critical function. Product decisions must transition from ideals to leveraging data and using intellectual rigor.”

Facebook was successful, Johns says, “because they evolved into making the right decision vs. the decision one person wanted to make.” The latter is often called the HIPPO, or highest-paid-person’s opinion. And it’s counter-productive in the growth phase. Founders must learn to let go, and executives must provide strong air cover to the growth team.

Johns notes that the support of VP of User Growth Chamath Palihapitiya was key to that evolution at Facebook.

2. Question everything

At Facebook,”Chamath was exceptional at [providing air cover],” says Johns. When needed Palihapitiya wasn’t afraid to argue. “He steamrolled when necessary to get the right thing done for Facebook. He was very good at that.” Johns means this as a high compliment to Palihapitiya and the rest of the Facebook’s executives, who grew to trust the growth team as it steadily improved the company’s metrics. Palihapitiya himself praised Mark Zuckerberg’s ability to prevent anyone’s ego, including his own, from getting in the way of the company’s growth.

“Avoid a get-off-my-lawn mentality,” adds Johns. “The way things usually work is that the management team goes to the product team and asks them to deliver on X.” He suggests growth companies reverse the direction, with horizontal growth teams providing viewpoints and evidence to the management team, who should then apply the correct lens to the insights.

Johns notes: “At this stage, the growth team is a strategic advisor to the company.”

3. Executives must continue to focus

When he left Facebook for Twitter, Johns discovered what happens to a growth team without a strong executive sponsor: It was a much harder process to transition into Miner mode from Prospector mode, and the company was having a hard time expanding to international markets. So the team had to start small, working skunkworks-style in near secret. But once it found those 70,000 users a day mentioned in my earlier post, things changed. “Once we proved the value of taking a different approach to growth, on at least a subset of the data, we were given more trust and we were able to help grow the company.”

4. Watch Churn

Both Johns and Ellis pay attention to churn. At scale, a failure to watch churn metrics, especially once spending begins on paid search, or awareness, can rapidly lead to an upside down cash flow problem and a subsequent death spiral. HomeJoy is the most recent example of this. At the early stage of a company, it’s not a metric to optimize on, but it is a metric to know, because a high churn rate means product fit may not really exist.

5. Build structure and scalability into process

You’ll hear Johns — and every great growth marketer — obsess about the importance of rigourous scientific method approaches to Miner-style growth hacking. You’ll hear Ellis talk about supporting even early-stage companies with some kind of discipline and focus about what efforts will get done.

When I worked at eBay in 2003-2005, one of the key lessons was “What you decide not to do is just as important as what you will do.” Meg Whitman and the executive team in full hyper growth mode had a clear process for bubbling up strategic ideas, distilling them based on data, and communicating them back down to the company in short, prioritized annual strategic initiatives. There were never more than six points on the list for the entire company. Employees, and there were 6,000 of them at the time, were instructed to inform their managers immediately if any objective assigned to them did not fit in those strategic objectives. That executional rigor is part of the legacy of good tech companies.

Growth hacking is a part of marketing, and it’s here to stay. The only question is how it will manifest in your organization. Match your growth phase to your marketer’s mindset and the chances of success increase. Whether you’re prospecting or mining, gold is out there, just waiting to be discovered.

Are you a Prospector or a Miner? Let’s continue this conversation. I hope to meet some of you at the GrowthBeat conference next week and hear about your experiences. Or you can reach out to me on Twitter: @mjb_sf.

Special thanks to Shane Johnson for his help in finding the miners and prospectors metaphor for this growth hacking framework.

Melinda Byerley is founder of TimeShareCMO, where she sells picks and shovels to growth Miners and Prospectors. She was previously a Miner-style growth hacker at eBay, PayPal, Linden Lab(Second Life), Check Point Software and Poll Everywhere. She has been a Prospector-style growth hacker for nearly a dozen startups including her last venture and current clients. You can follow her on Twitter: @mjb_sf.

VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Learn More