One of the more common questions we get asked is why a startup needs a growth team and not just a marketing team. The assumption behind the question is that marketing and product are different teams.

But in our opinion, marketing and product have to be intricately tied together, and “growth” is the term we use to describe this relationship. A growth team combines engineering, product, and marketing, working in concert in a cadence of rapid, data-driven iteration to not just acquire customer but keep them coming back and telling their friends.

Our process is heavily in debt to Dave McClure who developed the AARRR framework, which organizes product development as a series of experiments to improve one of five levers:

• Acquisition: how many people come to the site.
• Activation: how many people sign up and have a good first-time experience.
• Retention: how many people come back.
• Revenue: how many people are monetized (directly or indirectly).
• Referral: how many people tell their friends.

Growth is not tactics

There are dozens of strategies and tactics one can employ to manipulate these levers. But the key to success in growth is process, not tactics. What worked for other companies might not work for you. You’re talking to different types of customers, potentially on different mediums, at different times.

Additionally, channels and tactics constantly change. The effectiveness of an approach tends to diminish over time as it becomes ubiquitous.

So what’s needed is a process that isn’t dependent on any specific channel or tactic — one that can reliably generate consistent wins for your team over time. The process we use is heavily influenced by Sean Ellis and Brian Balfour, and includes the following steps.

Step One: Validate Product Market-Fit

Are you making something that enough people want? If not, you’re wasting your time.

It is definitely possible to find tactical wins at the acquisition level. But if the users you acquire don’t stick around or turn into revenue, you’re effectively lighting money on fire. It’s almost impossible to grow a product at scale without product-market fit.

P-M fit is fuzzy, but is best identified by monitoring the engine of your product — the activation, retention and referral metrics. Solid activation means your product’s promise resonates with customers. Retention means it delivers on the promise. And referral means the promise is great enough that people are compelled to tell others about it.

Until you’ve reached this point, scaling does not make sense. In fact, according to the Startup Genome project it can kill you — fully 70% of startup failures have been attributed to premature scaling.

Step Two: Hypothesize Models for Growth

You don’t know what will work until you start testing. So it’s important to identify as many potential strategies as possible. For us, this involves looking at each stage of the funnel and hypothesizing as many potential approaches as possible for influencing them.

Some ideas are straightforward, some are unique, others are weird and probably won’t work. But you want to capture all of them at the outset. You will prioritize and filter in the next step.

While it’s important to not treat what others have done as prescriptive, that doesn’t mean you should ignore what’s worked for others. In fact, a huge source of inspiration for tests you can run will come from knowledge of strategies others have leveraged successfully.

Step Three: Prioritize the Backlog

Once you’ve built up a list of hypotheses, you prioritize the backlog using a system called ICE scoring. ICE stands for Impact, Confidence, and Ease. For each test idea you ask three questions:

• “How likely is this test to have a great impact on growth?”
• “How confident are we that this idea will work?”
• “How easy will it be to run this test?”

Each factor receives 3 scores from 1 to 10. The higher the ICE score, the sooner you want to test it.

Step Four: Test. Learn. Double Down. Repeat.

From here, it’s a matter of rapidly testing ideas, taking care to document learnings along the way. This is crucial, especially when you reach a high volume of weekly tests — your learnings are used to inform future tests, and help to avoid making the same mistakes twice.

Just as The Lean Startup advocates for rapid product iteration, the growth process prioritizes speed through the build, measure, learn loop. The team that wins is not the one that has the best tactics, but the one that discovers the right tests sooner.

As you find wins, you want to double down on what works. We advocate spending 80% of your time optimizing and building upon winning tests, and 20% testing new approaches.

The best way to achieve consistent growth is to achieve a consistent velocity of testing. Brian Balfour calls this “a growth machine.” You need a machine in order to consistently scale channels that work, while allocating time and budget toward testing new channels.

Because “low-hanging fruit” is a falsity — you can’t know what those things are until you’ve tested multiple channels, and established a process for discovering new wins. Even if you do stumble across an “easy win” on accident, you won’t be able to scale that until you have established a process for discovering new wins. We usually follow the 80/20 rule here. Spend 80% of your time doubling down on what works, and 20% testing new channels.

Growth Is Optional

Most companies want to grow, intend to grow, staff to grow. But far too often they fail to engage in a rigorous process of experimentation, leveraging a cross-functional team focusing on all five phases of the customer experience. By developing a process for growth inside your organization you maximize your product’s long-term chances for success.