7 Keys to Starting A Successful Growth Team

There’s a lot of commotion around the importance of a growth team, but what steps are involved in successfully establishing one? It’s a frequent discussion I’ve had with friends in growth, so I thought it’d be helpful to outline my seven steps to starting a growth team.


  1. Company Commitment

Growth isn’t a role you simply hire for and expect to drive results independently. To optimize growth, it involves input from all departments and is a strategy the entire company is aligned with. Because what role or department doesn’t want the company to grow?!

What type of hire should it be?

It’s important to be realistic though when starting a growth team. Hiring a lead Growth role is the first step. This person should be knowledgeable in the suite of modern marketing tools, yet has engineering experience and can effectively work with product. Scrappiness and ability to define an actionable strategy are two essential traits for this role.

One of the biggest questions is whether growth is a part of marketing, product, sales, or engineering… The best answer is all of the above. Growth requires approval from design to A/B test a pop-up box for content pages. It involves social media to execute a twitter ads campaign. Developing a referral system needs input and potential development from engineering. The fact is growth sits closely between product and marketing

Growth teams that are highly technical and include engineers can move faster, but this is not a requirement as long as part-time engineering resources can be allocated for more technical tasks.


  1. One Simple Metric

Boiling down your startup’s definition of growth into one simple metric the entire company can understand will help streamline your growth strategy. This activity isn’t difficult, but requires your growth manager to understand your business model and offering at the deepest level.

For example, should your metric be the # of Monthly Active Users (a metric utilized at Pinterest) or the # of leads your company is generating? These sort of metrics provide insight, but which one is the most appropriate definition of growth for your company? Experiments will be designed based on this metric, which means you will want to keep it constant so you can make actionable decisions based on these results (I recommend committing to a metric for at least 2 quarters). Note that it’s important to have secondary metrics to help with analysis, but when forced to make a decision, being able to point to a source-of-truth metric will streamline this process. This allows your growth team to easily prioritize tasks and set goals when increasing and optimizing for this metric (see next point).

One metric also allows the entire company to quickly understand how their work affects growth. At the end of the day, all departments can positively impact this metric since it’s all encompassing. This is powerful, as it can align the entire team to contribute and help increase growth.


  1. Establish a growth experiment list

A growth experiment list is the list of all the experiments your company is interested in running. Experiment examples include an A/B test (copy, emails, design, etc), new product features, separate growth projects, or any other changes/updates that can be measured. A list is essential because it allows a growth manager to prioritize which experiments to set up and execute based on its impact potential, chance of success, and effort required. I like to use Sean Ellis’ ICE process to help with the prioritization process.


  • ICE stands for:
    • Impact potential. What sort of results could this test provide?
    • Confidence in its success. What are the chances the experiment will succeed?
    • Effort required. Will an engineer be required to build out this test? How much time will it take?


  1. Execute and develop a process

It’s now time to put those prioritized experiments into play. Proposing what “success” means before execution is helpful for reference later. Have your team aim to have a certain number of experiments pushed out each week (3 is a good number to start).

A scrappy, get things done growth manager and team are essential since growth involves a variety of marketing tools and skills. For more technical experiments, this is where allocating engineering resources, whether it’s part-time for a small startup or a full-time headcount for a larger company, comes in.

It’s important to utilize ICE as a source-of-truth when reviewing whether an experiment was successful. Or not. Oftentimes, there won’t be a clear cut winner, which is why the agreed upon metric and definition of success should be used as a reference.

Implement the improved experiments, push more tests, and repeat! Keeping track of what has been run is key, and provides clarity over what strategies have been tested.


  1. Allow the entire company to participate

Every role at your company has expertise within their domain on how to increase growth, which should not be ignored. It’s also natural for one person or a small team to eventually run out of great experiment ideas. By allowing anyone to propose growth experiments, you enable everyone to be a part of growth, which is a key to developing hypotheses that increase the quality of experiments on your list.

Build a process to enable participation by the team. Examples include setting up an experiments@ email address or google form that allows an employee to submit an experiment and its proposed ICE. Communicating growth updates to the team should be welcome, whether via a scheduled report or update session.

With that said, the growth team is the owner of the experiment list, which should be clear when setting expectations with the company (similar to a product roadmap). Opening growth ideas to the whole team can become a distraction when not managed properly, which is why keeping metrics, strategies, and goals simple and transparent are important.


  1. Understand that failure is part of Growth

You cannot expect all experiments to be successful. Doing so would stifle your team, and limit the amount of “high impact” growth experiments your team goes after, which occasionally lead to significant results conservative tests cannot achieve. It’s similar to a batting average. If you only go for conservative singles, you’ll never get that home run experiment that provides a jump in growth.

Cat Lee, the Head of Growth at Pinterest, has her team define tests two basic categories. 70% of their experiments affect the core product and are of medium ICE levels, while the remaining 30% are “moonshots”. Moonshot tests may be of a lower confidence level and require increased effort, but could provide drastic increases in growth, warranting their development. A healthy balance of experiments should be emphasized by the growth team, which help provide diverse wins of varying impact.


  1. Grow your team!

If your growth team is proving its ability to accelerate the trajectory of your startup, work on providing additional resources to your team. It’s important to keep the cadence of growth initiatives consistent, and increase resources when they’ve proven successful. The quicker your team can move, the sooner you’ll receive results. A growth team should consist of PMs, engineering, and marketing-related roles that can execute non-technical experiments.

Good luck!