How to Start a Growth Team
Lessons learned from starting the Yahoo Growth & Emerging Products Team
As 2012 drew to a close, a big decision was stressing us out. Our small team of five had been working on a nascent product called OnTheAir with the mission of making the most interesting people in the world more accessible. After two years of pursuit, we decided that our product wasn’t going to work, and that it was time to sell the company. Because our options at the time were primarily talent acquisitions, our main goal was to find a high impact project that the team would be excited about. Yahoo gave the most compelling pitch: join the soon-to-be-announced Mobile and Emerging Products group led by Adam Cahan, and help Yahoo build a mobile presence from scratch. The hard decision wasn’t whether or not to join Yahoo, but what to do once we got there.
Thankfully we were given tremendous freedom to carve out our role in the newly formed group, and we made impact our number one priority. While a number of exciting new product teams pitched us to join their efforts, we had an inkling that we could accomplish more at Yahoo if our work could be leveraged across all of Yahoo products. Moreover, we craved to work on something at a large scale after working on a pre-product/market fit product for two years. Starting to think harder about the idea of a Growth Team, we reached out to a number of Growth leaders to hear about their experiences — people like Hiten Shah, Greg Tseng, Andrew Chen, Othman Laraki, and Elliot Shmukler. These meetings were the turning point for us, as we came away impressed by both the individual acumen of these folks and the massive impact they were able to have with relatively small teams. So despite zero experience working on growth-related projects, we decided to start the first ever Growth Team at Yahoo, with a focus on growing Yahoo’s mobile products.
Since then we’ve come a long way. Yahoo Mobile as a whole has been greatly successful, growing from essentially non-existent revenue to more than $1B in revenue in 2014. While I’m not allowed to share specific numbers externally, our team’s initiatives have added substantially to both the revenue and active user counts. More importantly, the team keeps expanding with great people. Last year we were thrown an incredible stroke of good fortune, inheriting one of the very best Growth leaders in Arjun Sethi — who now oversees the Growth Team — and a whole team of growth minded folks from the MessageMe acquisition. As an aside, I suggest you check out Arjun’s killer consumer product framework that we use for thinking about consumer products.
Two and a half years in, we finally feel like we have a team that’s operating at scale and kicking butt. While we still have A LOT to prove, now is a good time to reflect on our learnings around starting a Growth Team from scratch. We hope that our successes and failures can help guide you as you pursue growth at your own companies. To start a Growth Team, you must address the following questions…
- Is your product ready for growth?
- Who will be your executive champion?
- How should you structure the team?
- How should you spend your first 30 days?
Is your product ready for growth?
Growth has become a bit of a catch-all term, and the term “growth hacking” has its place among some of the most overloaded and gimmicky catch phrases in tech. Thus I’d like to offer up a definition for what a Growth Team actually does, largely borrowed from Arjun: Growth Teams expand the value of, and customer base for a product that is already working through rapid, data-informed experimentation. The corollary to this definition is that you should not start working on growth unless your product has product/market fit.
I won’t go into how you know if you have product/market fit as this topic is well covered by other writers. If you haven’t read up on this, I recommend this post by Brian Balfour, this talk by Hiten Shah, and of course the seminal piece on product/market fit by Marc Andreessen.
Above all else, this is the one thing you must be honest about. We were given this advice early on and ignored it several times, trying to apply growth methodologies to products and behaviors that were not yet established. Every time we tried this we failed, and you will fail too if you don’t take this advice to heart.
Who will be your executive champion?
So you have product/market fit and you want to get the Growth Team going. Who do you need to convince?
Andy Johns first wrote about the importance of executive sponsorship for growth initiatives, and we have experienced this to be critical as well. My recommendation is that the product lead of the Growth Team report to someone on the executive team, ideally the CEO/Founder. The engineering lead for the Growth Team should report to the engineering counterpart of your executive sponsor.
The Yahoo Growth Team sponsor was Adam Cahan, an SVP at Yahoo who ran all of Yahoo Mobile at the time. Adam championed our team and evangelized our wins. His support was (and still is) one of the most important factors in our success. Here’s why getting an executive champion is so important:
You’re going to work on someone else’s product
The product you’re planning to grow already has a product and engineering team, a codebase, a team culture, and often a set of priorities that have nothing to do with your own. You are going to impact all of those things when you insert yourself into their process.
When we scaled out our service to cross-promote apps to Yahoo mobile web users, we had to write a service that processed every single mobile web request at Yahoo. It sat at the top of the user experience for every mobile web property. Product owners across the company were understandably nervous about this, and having a strong mandate to do this made it possible. Even with the mandate, we still had to do a lot of work to get support from the partner teams — it was vital that we respected the host team’s culture, integrated into their workflow/schedule, and clearly articulated the importance of the Growth Team’s work. However, without a mandate, the project would not have gotten past Go.
You’re going to try controversial things
When we first started down the path of app promotion, we initially got strong push-back from folks in the brand team around some of our experiments that deviated from the company style guide and marketing voice. To be fair, the brand team was doing exactly what they were supposed to be doing, and only Adam’s clearance allowed us to experiment freely. While I can’t share specifics, I can say that some of our best performing variants were a bit unconventional. The freedom to experiment with unconventional methods is essential if you want to grow, and you need executive air cover to do this.
How should you structure your team?
First, the easy part: a Growth Team is just a product team. That means you need awesome engineers, designers, and PMs. Depending on what you work on, you probably want some people that are good at marketing as well, although often that falls on your designers/PMs. One trap I frequently hear people fall into is that they empower a PM to work on growth initiatives with no design or engineering team. The idea is that this person will learn about growth and push for it to get on other teams’ roadmaps. This almost never works, as your priorities will always play second fiddle to the lead PM on the host team.
A more interesting question deals with the personality traits you want on your team. Andy Johns states that at Facebook, Growth Team members are data-driven and aggressive, with an appetite for risk. Brian Balfour looks for people with a relentless drive for learning and improvement for HubSpot’s Growth Team. Chamath Palihapitiya talks about how he looks for people that care about delivering value over perception in one of the best ever talks about Growth.
At Yahoo, we have found three personality traits to be particularly important for members of the Growth Team:
Open-minded
This is the #1 characteristic we look for because the best ideas are often counter-intuitive, and you get the most leverage out of growth experimentation when you can test vastly different theories. If you are overly dogmatic on what will work and what won’t, you will likely miss some of the biggest wins. This is doubly important for Growth Team leaders (PMs, Tech Leads) who have the power to kill ideas. One of the first things Arjun said to me when we met was,“I can’t give you any meaningful advice about your product specifically, because every situation is different, and you’ll have to find what works. I just have a process that I believe in that I can share with you.” This willingness to be open minded is exactly what you want in a growth leader because what works is constantly changing: the growth strategy for one product might not work for yours, the best channels six months ago are likely saturated, and your customers have a unique intent when coming to your product.
Thinks for herself
In the same Chamath talk, he states that the most important cultural achievement of the Facebook Growth Team was to “eliminate ego and invalidate lore.” All companies have false narratives that propogate, often by high ranking people. You want people who are inherently skeptical of what is accepted as “Truth” at your company, and who are willing to form their own opinions based on data and then test them. Often times these lead to the most important findings and biggest wins. For example, I believed late last year that we had exhausted all of the growth/optimization from our app promotion system. One of my new teammates from MessageMe, Sergei Sorokin, disagreed openly and then proceeded to execute on a plan that doubled the amount of app installs we were driving in the next quarter. If Sergei hadn’t been inherently skeptical of my position, which was informed by living and breathing this same project for a year(!), we would have left a lot of growth on the table.
Quickly learns new things
The engineers on your team will write server, web, and mobile app code. Your PMs will need to be equally comfortable diving deep into data and writing marketing copy, especially early on. The marketing tactics you try constantly change and range from paid acquisition to SEO to notifications/re-engagement. Almost no one is deep in all of these areas, and you need people that can quickly pick up new skills. The person that is the world’s leading expert in Android/machine learning/Systems/Whatever, but has never written a website or mobile app, is probably wrong for your team.
We of course look for all of the usual traits you want in a team member: excellence at her craft, driven, team-first mentality, fun to be around, etc., but we carefully screen for the three traits above in all of our hiring. For what it’s worth, people that have been founders and started companies often seem to exhibit these characteristics.
How should you spend your first 30 days?
So you’ve got a killer product team and an executive sponsor. Awesome. You probably also have some doubters in the organization. Some of the common objections include “product teams should be accountable for user growth, why do we need a growth team?” and “we already have a marketing team!” Moreover, there are lots of shiny new features on the company roadmap that would really benefit from having your engineers. The pressure is on to prove your worth, so where should you start?
There are three things you should do in your first month:
Focus the team around one metric
Great growth teams usually focus exclusively on one metric. Note that this is different for every company — Wealthfront focuses on dollars under management, AirBnB focuses on nights booked, Linkedin focuses on new user signups, and social media properties tend to focus on DAUs. That said, they all have a single metric to focus on because it makes prioritization of growth projects much simpler, and it focuses the team in a way that allows everyone to contribute ideas. Josh Elman has a good deck on how to think about what your single metric should be here.
Generate a few quick, high-impact wins and evangelize them
Remember the doubters? This is the easiest way to get them on your side. I recommend picking a flow that has a ton of traffic and that hasn’t been rigorously tested and improve it dramatically. New user signup and onboarding is usually a good place to start. More importantly, pick something that you can change quickly, meaning you can show impact in just a couple of weeks of work. There is almost always low hanging fruit you can pick by running tests on copy, calls to action, or subtle UX changes. Once you see results, create a case study and promote it widely throughout the company. This will cause other teams at the company to want to start working with you, and will ensure the ongoing support that you need.
Start investing deeply in the tools for analytics & experimentation
Once you’ve hacked your way to a few quick wins, you’ll likely discover that the analytics and experimentation tools you’re using won’t suit your long term goals around deep analysis and rapid experimentation. Growth Teams tend to win by creating compound interest, meaning lots of small/medium wins rather than a few big ones. This means you need to optimize your team for high experiment throughput, and nothing will slow you down more than weak data/experimentation tools. Every company’s situation is different, and the best tools change every quarter, so trying to recommend an exact toolset isn’t helpful. That said, make sure you reach out to people that have done this well recently and bring the best-in-class tools in-house.
In summary…
If you want to start a Growth Team, we recommend that you…
- Ensure you have product/market fit before starting.
- Find an executive champion.
- Put together a product team with people that are open minded, that think independently, and are capable of quickly picking up new skills.
- Knock your first month out of the park by selecting one key metric, scoring/evangelizing some quick wins against that metric, and starting your investment in experimentation/analytics infrastructure.
Now go crush it!
Thanks to Ana Braskamp and Hiten Shah for reading drafts of this post. Thanks to Abel Allison, Dan Hopkins, and Mike Kerzhner for going through this journey together from the start.
