Sean Ellis spent several years working on short-term projects with startups, where he would lay the groundwork for growth and try to 'stack the odds' for that startup to succeed. When he started to look at the profiles of VP Marketing people at various startups, he didn't have a lot of confidence in their abilities to really drive growth. Talking about his experiences and this gap in marketing ability inside of companies, he and some fellow tech people coined the term 'growth hacker' to define the type of person that can really drive success through growth.
At the Growth Hacker Conference last week, Sean spoke about what he's learned working at Dropbox, Xobni, Socialcast, and Eventbrite (among others). He's pulled all of his learnings into what he calls the 'Growth Pyramid', which has product/market fit as its base, followed by a stage called 'transition to growth', and topped with 'growth'.
Product/market fit - The only thing that matters
Sean mentioned the writings of pg and pmarca as those that best confirm and solidify exactly what PMFit is, and help to stress its importance. If you can get to PMFit, 'you're no longer in mission impossible… it's now mission frustrating'. Having seen this at Dropbox, once you know the formula is right, growth is really about simply sprinkling on some accelerants - but the product was really working at at PMFit first.
To validate PMFit, Sean has a simple test that he runs - if 40% or more of your users would be disappointed if your product no longer exists, you're at PMFit. Having worked with companies that had a lower percentage response rate to the question, Sean confirmed that those types of companies are difficult to grow. But once you have that type of enthusiastic base you can start to think about transitioning to growth. Keep in mind that PMFit is perishable, so speed and getting to the next step (and getting whatever resources are required for that step) is very important. Also, Sean's personal view is that the 60% that are perfectly happy without your product are not worth your time - don't chase those people, and instead worry about the people that you can actually gratify - double down on their feature requests, and pay particular attention to what they say and how they talk about your product.
Taking a bit of an aside, Sean spoke about the dangers of being an employee at a pre-PMFit company. He believes that there's not enough equity available to risk your track record at a company at that stage. He would actually run the 'how disappointed would you be…' test with the users of any startup that he was considering advising before he would accept the role. His experience has taught him and defined a 'playbook' for both the transition to growth and growth phases, and he knows that it's all about proper execution of certain tactics.
Key growth transition activities
The very, very first step once you have PMFit (if you haven't already) is to implement user analytics. It's critical to get an understanding of where people are coming from, where in your process they are, if they're converting, etc. With this information, you're in a much better position to continue the next steps on your way to growth. Secondly, you have to clearly identify the 'must have' benefit of the product. Figure out why your users would be disappointed if you vanished, and understand what primary benefit your product offers. This will then allow you to map those benefits to effective hooks. Next, you need to start thinking about how to optimize your business model, and begin experimenting with different approaches - keeping in mind that a short-term loss may be required for ultimate long-term gains.
Sean gave an interesting example of a company that he worked with. The product was good, and they had started trying to grow… but it wasn't working. They experimented a lot, but couldn't get their growth curve to hit any sort of inflection point. Eventually they decided to sit down and fully map out their funnel. They diagnosed areas of friction, and a/b tested the entire funnel (landing pages, messaging, trial process, installers, etc.). They had their own theories of why users weren't converting, but none of the a/b tests confirmed their assumptions. Finally, the team decided to insert qualitative feedback from their users at the various stages within the funnel. At the point where a lot of users were dropping off (the 'install' page), they asked 'What's preventing you from downloading?'. It turned out that users didn't actually believe that the product was free. The team now had a clear answer on what to improve - adjusted landing pages that contrasted paid vs. free immediately cleared the bottleneck, and negated any theories the team had about the impossibility of downloadable software to experience hyper-growth, the need for a different installer, etc.
Finally, in the transition to the growth phase it's important to remember that: Growth = Desire - Friction
Sean's advice is to focus on driving momentum while also reducing friction. The common approach to this is around optimizing your funnel and a/b testing - but Sean also warned agains being overly aggressive with this style of testing. The risk is that eventually the message starts to not relate to the product as much. You maybe increase your signup rate, but if you're driving momentum that's not related to real value, who cares? Users that come in through optimized but inauthentic means will eventually realize that you're not offering the value that they converted for. Sean's final tip for this stage is to really focus on 'must have' users, people that can't live without your product. Dig deep into what the benefit is that they're getting and focus on bringing that messaging into your hook.
Now, time to grow
Once you've done all of that optimization and transition to growth, you're now in a position to actually grow! The friction has been removed, your product is a 'must have' for a certain group of people, and you really understand the benefit of your product that's driving momentum. Keeping in mind that growth is never easy, Sean did offer some thoughts on how to establish the right culture and team to facilitate this stage.
When it comes to culture, Sean mentioned that when he left Dropbox the entire team really understood data and the different levers that they had control over. At a high level, everyone at the company was a growth hacker. If people have the support of a team that believes ultimately in growth, someone who comes in to manage that goal will be able to really push and achieve great things. In terms of team, Sean's ideal composition is designer + copywriter + data person (+ ideally a developer). This team needs to be able to observe, and get inspired as to what they can take advantage of - both directly from your product as well as being inspired by what others are doing. He added that you can rarely knock something off directly and it'll work. Instead, you need to get to the same intention. At the end of the day, a growth team needs luck, inspiration, responsiveness, and tenacity to succeed.
Wait - what about networks?
Sean added a final caveat for companies that have an inherent network effect (e.g. AirBnB, LinkedIn, Facebook). These are the types of companies where the value to each individual users goes up as more users are added to the network. In this case you can't say you'll build value and then get users… the value of the product is itself a function of the suers being there. This type of value is something that you can't really stage either. His advice is that that growth for these types of products is much more fluid, and you need to be able to optimize on the fly, and measure changes over time as the network changes. Over time the value might be related to the community that develops, or access to group knowledge - but to keep in mind that things may change over time and as you optimize for growth.