We interviewed 20 growth leads at some of the fastest-growing companies like Pinterest and FanDuel to help you answer this crucial question.
After more than 30 hours of interviews, one thing became abundantly clear: choosing the right growth team model not only unlocks growth, but also strengthens culture. (click to tweet)
Two popular models for structuring a growth team emerged from our interviews.
One is the Independent Model; the other is the Functional Model. Both are good starting points for how you may want to structure your growth team “when the name of the game is up and to the right” (Zack Onisko, Head of Growth at CreativeMarket). Each model focuses on “developing a deeper understanding of your pattern of growth” (Mike Duboe, Head of Growth at Tilt), and both focus on unlocking growth throughout the entire funnel.
Uber and Facebook use the Independent Model. Two versions of this model that we most often came across in our interviews are organized as follows:.
In each of these two versions, a VP of Growth heads the team. At Uber, for example, Ed Baker is the VP of Growth and reports to Travis Kalanick the CEO. Ed leads 100+ people with product, marketing, engineering, design, and data expertise in working on both demand and supply-side growth.
Also according to our contributors, Uber uses this model to build strong team DNA around speed and iteration. Uber values speed and iteration because they enable the Growth team to build strong feedback loops into Uber’s product.
Feedback loops help a cohort of users create another cohort of users, and are often triggered by search, paid, and referral mechanisms. For example, riders who take their first ride based upon recommendations of existing riders are the result of a referral feedback loop. Drivers who become drivers by clicking on “Drive With Uber” ads are the result of a paid ad loop. In each case, these feedback loops help existing users create new users, whether on the demand or supply side of the product.
Uber’s Growth team’s job is to make these feedback loops as powerful as possible in order to increase their multiplier effects. For instance, say the multiplier effect on Uber’s referral feedback loop is “10x.” What this means is that the number of riders acquired through linear channels (e.g. paid ads) will be 10 times greater. If the team acquires 1,000 riders through a paid ad, these new riders will create another 10,000 riders through the referral loop.
The greater the multiplier effects, the faster Uber can grow.
Tensions and Resolutions (Independent Model)
Casey Winters, the Lead Product Manager of Growth at Pinterest, said that the model is not without its tensions.
When Pinterest used the Independent Model, the growth team’s numbers kept going up. In fact, Pinterest’s metrics grew so quickly that other teams became concerned growth was coming at the expense of the broader user experience.
For example, when the Growth team implemented a new sign up flow, it drove a significant increase in activations, as well as concerns. Other teams thought that the flow provided for a poor user experience. Data, however, indicated otherwise, despite the counterintuitiveness to these other teams.
The tension between maintaining user experience and attaining powerful growth metrics is the greatest challenge our contributors said they face in leading independent-growth teams. (click to tweet)
Is growth coming at the expense of user experience is the argument that is heard time and again, and if not countered appropriately, “it can create distrust within the organization.” This is according to Winters, who went onto say that “people should know (and trust) that their growth team is creating the best experience for users.”
How can you establish this fundamental trust across teams?
One good way to establish this trust that Casey refers to can come from having in place a written agreement that says your growth team will not try things that compromise user experience for growth metrics. This comes from another contributor who suggests making your CEO the owner of this document because a company culture starts with the CEO as well as having your VPs of Growth and Product help write the agreement. This way you can reduce another tension that this contributor sees in the Independent Model.
Growth initiatives will inevitably overlap with product initiatives, according to Naomi Ionita, Head of Growth at Invoice2Go (previously Head of Growth at Evernote). Based on her experiences leading these two independent growth teams, she explains that if your roles and responsibilities are clearly defined and broadly communicated, or your Heads of Growth and Product have a strong working relationship, this tension is avoidable. It comes down to “trust, transparency, and objectivity.”
According to our contributors, when people working on growth initiatives report to functional heads (e.g. Product, Engineering, etc.), other people believe that growth metrics will be achieved the right way. (click to tweet)
Pinterest, Twitter, LinkedIn, Dropbox, and BitTorrent use the Functional Model.
Three advantages make this model popular.
1. Functional heads (e.g. Product, Engineering, etc.) decide which growth initiatives get worked on
2. Functional heads balance growth initiatives against non-growth initiatives
3. The VP of Product is usually the functional head who owns growth
Let’s look at this last advantage in practice.
Pramod Sokke is the Head of Products for all clients on all BitTorrent platforms. He’s responsible for BitTorrrent’s growth metrics and non-growth metrics. Meaning Pramod must develop his product roadmaps for growth and non-growth KPIs because he’s responsible for both sets of KPIs. People not working on growth initiatives at BitTorrent are confident that he’s prioritizing growth initiatives in a balanced way to ensure the success of everyone in the long run.
Tensions and Resolutions (Functional Model)
Nonetheless, the tension between user experience and growth metrics still exists. It’s not as pronounced as it is in the Independent Model because of the advantages mentioned above. But still exists because a fair amount of “growth stuff” that works can be counterintuitive from a user engagement perspective.
This is why our contributors also advocate for longitudinal analysis. It answers the question of whether growth is coming at the expense of engagement by analyzing user behavior over time. Out of which comes a clear picture of whether a growth initiative not only acquires users, but also drives engagement.
We also learned that the Functional Model works best when the person who’s ultimately responsible for the growth metrics is data driven and recognizes that product is iterative and learning driven. It’s not about placing big bets on a product roadmap solely based upon gut or intuition.
For instance Pramod takes a very data driven and iterative approach to product. This makes Annabell Satterfield, the only growth PM on his team, far more effective. Why — because their two processes align well.
Annabell explained that her growth process is data-driven and iterative. It’s not about intuition or gut, rather about process: look at data (qualitative and quantitative), make hypotheses, push out tests and MVPs, look at the results, and rollout or start again with your new learning.
Now let’s suppose their processes didn’t align well to illustrate another potential tension in this model.
Say Pramod takes a non-iterative approach towards managing products Annabell is growing iteratively. Pramod would most likely ask Annabell for big bets for the product roadmaps and that would take 6–12 months to deliver. Annabell would give Pramod these big bets, but she’s likely to invalidate them within weeks (if not days) using her learning driven iterative process. This would only drive increases in product-planning costs, not revenue or users, according to Annabell.
Some of our contributors would resolve this tension by transforming the product roadmap into swim lanes. In the case of our hypothetical example, Pramod may want to incorporate two swim lanes versus only having one lane for all product work in each of his product roadmaps. In one lane would be big bets dictated by Pramod’s non-iterative approach. The other lane would have growth initiatives dictated by Annabell’s iterative process. Each lane would have its own dedicated product, engineering, and design resources to ensure that all work got done on time and on budget.
Leadership Buy In (“Must Have” in Both Models)
Even then, leadership would need to buy into a data driven and iterative mindset to drive growth. If leadership does not, you will fail no matter how you structure your growth team.
Growth is intertwined with product in the software world we live in — meaning it’s hard to know where product ends and growth begins. This is still a relatively new phenomenon that makes choosing the best growth team structure challenging.
In interviewing 20 growth leads, two popular models emerged: the Independent Model and Functional Model. Both are good starting points for how you may want to structure your growth team in the scalable growth phase of your business. This is when the name of the game is up and to the right.
The Independent Model lends itself to greater speed and iteration in running growth experiments. Nevertheless, it provides less transparency around growth initiatives that can create distrust within an organization. The Functional Model provides greater transparency by taking a more balanced approach toward growth. The tension between user experience and growth metrics still exists, however.
Interestingly, we didn’t find much difference in the eventual outcomes between the two models. Meaning one model is not necessarily better than the other for achieving a greater growth multiplier within a product.
This is a critical finding.
It suggests choosing a growth team model based upon cultural, organizational, and strategic fit may be the better path forward. Don’t select for growth. Select for fit.
Pick the model with the tensions that your ‘team DNA’ can best resolve today, and choose a growth team leader with attributes that also fit the situation. (click to tweet)
How well your growth team leader interacts with your product team will make a tremendous difference in the success of the model you choose, and in your company as a whole.
Andrew (Drew) McInnes is at Tradecraft. Before Tradecraft, Drew was a Product Manager of Growth in social and mobile games (most notably at Zynga) and wants to continue to help build growth engines . So if you’re looking for a product driven growth person, feel free to contact him on LinkedIn. If you’d simply like to be friends, connect with him @ahmcinnes, or email him at “mcinnean (at) gmail.com”.
Daisuke (“Dice-K”) Miyoshi is a Growth Marketer at Tradecraft / ex-Googler who is passionate about helping mission-driven startups build their growth engines. If you’re looking for a Growth Marketer with paid acquisition and data analysis experience, feel free to contact him on LinkedIn. If you’d simply like to be friends, connect with him @dicek344 or email him at “daisuke (at) tradecrafted.com”.
Their insightful contributions made all of us better growth-team leaders.
Published in #SWLH (Startups, Wanderlust, and Life Hacking)