Reading marketing books can be a very disappointing activity. The time it takes a title to travel all the way from the author’s head to the reader’s hands is often longer than the half-life of marketing trends. Digital marketing changes by the second and books just can’t keep up.
But still, sometimes an author does their job well and describes something long-term. That’s what Sean Ellis and Morgan Brown have done in Hacking Growth: How Today’s Fastest-Growing Companies Drive Breakout Success.
The book explains how to build an experiments-driven culture and how to take decisions based on data. It is valuable to people who are still not familiar with growth hacking (and are not tired of hearing how Dropbox did everything right :)).
Building your Growth team
The first part of Hacking Growth goes into the details around developing your experiments team and structuring its work. In the ideal case, you’d need:
- A growth lead who manages the process, leads the meetings, moderates discussions and functions as a project manager. It’s best for that person to have a full overview of the customer funnel. It isn’t necessary for them to be part of a specific department if they have good organization skills.
- A product manager or a person with deep level product understanding.
- A developer who can say if an experiment is technically possible – and who will be charged with putting it into practice.
- A customer success manager or a person who’s gathering customer feedback. They’d know how we the customer experience can be improved.
- A marketing specialist. That can be a person charged with a specific step of the user funnel or a particular channel. The participant can change depending on the current cycle’s focus.
- A designer. They will take less time working on the implementation if they are part of the planning.
- A data analyst, who can take out the necessary metrics and influence the way an experiment is set up.
If you feel like this is a huge investment of human capital – you are totally right! To make it leaner, at the beginning, one person can take on several different roles. Say, a product manager can be the growth lead and also be responsible for the development work. To limit the risk, you can limit the scope of the experiments. For example, you could narrow the focus to a particular metric or channel. The important bit is to start and set up a regular process of ideation, testing and measurement.Growth cannot be a side project.Click To Tweet
An important note on the setup is giving enough freedom to the growth team to do their job. The authors put it very well:
Growth cannot be a side project. Without clear and forceful commitment from leadership, growth teams will find themselves battling bureaucracy, turf wars, inefficiency, and inertia.
High tempo testing
The aim of the growth team is to set up an ongoing testing process. What we mean by “tests” is a series of changes that have a measurable impact on results and return data within 2 weeks. Why precisely 2 weeks? It’s because the regular iterations help us get feedback from users and quickly learn from the market. No matter how useful or interesting a test may be, if it takes 2 months to complete, it will bring little value. You can read more about the reasoning behind that in this blog post by Growth Tribe.Conversion rate = Desire - FrictionClick To Tweet
In Hacking Growth, there’s a cool idea on thinking about conversion rate. Basically, the conversion rate is the result of the customer’s desire to try out your product minus the friction of actually trying it out. The equation above is a neat way to think about conversion rate. We focus too often on the first part, the desire of potential users to test our product. The more effective tactic is to remove customer friction. This is the end goal of regular testing.
Every person in the growth team participates in a weekly meeting. The goal here is to review the results of the latest tests and generate new ideas. Those ideas are scored by different criteria, and the most common scoring is the ICE format. ICE stands for:
- Impact – if the experiment succeeds, how big of an effect will it have on the key metric? We’re looking for statistically significant results.
- Confidence – how sure are we that the experiment will bring successful results?
- Ease – how easy is it to fulfill the experiment and how big of a monetary or time commitment does it require?
To shorten your meetings, every team member describes and scores their ideas before the meeting. Use the ICE score as a leading indicator of experiment feasibility, not the end-all and be-all of prioritization. You can always choose another experiment off the list, even if it doesn’t have the highest score. In this way, you will cut long discussions whether a specific idea is a 5 or a 6 in Ease. The team will focus on the actual prioritization and further planning.
Your product is the key to growth
One important thing worth mentioning – there’s a Step 0 in growth hacking and that’s making sure you have the right product to experiment with. Does anyone care about what you have to offer at all? Too few people test their product-market fit.
Sean Ellis has a trademark way of testing that. He uses one question only: How disappointed would you be if this product no longer existed tomorrow? The possible answers are „very disappointed“, „somewhat disappointed“, „not disappointed“ and „I no longer use the product“. According to Ellis, 40% of very disappointed users means you have something worth the growth efforts.
There’s a series of other questions you might want to ask:
- What would you use as an alternative to the product if it was no longer available?
- What is the primary benefit that you have received?
- Have you recommended it to anyone? If yes, how would you describe it?
- What type of person do you think would benefit most from the product?
- How can we improve it to better meet your needs?
When you’re sure you have a viable product, you need to identify the primary goal and the elements that influence that goal. Here’s an example. If you’re running an online store and your primary goal may be to generate high revenue. This will be influenced by the number of website visitors, the conversion rate, the average order value, and the repeat purchase rate. Then, your growth experiments need to focus on improving any one of those key elements.
Ellis and Brown give many ideas on what to experiment with at every stage of the growth funnel. They cover separately acquisition, activation, monetization, and retention. The best bit here is the real-life examples. They will give you ideas and will also serve as inspiration for your own growth process.
I recommend Hacking Growth and I suggest you don’t postpone and read it now. Not because it’ll be less relevant in a couple of months, but because you need to start experimenting as early as possible. If you don’t want to take my word for it, check out some other reviews of the book.
Here’s a nice warm-up: a talk by Sean Ellis. It covers the main points of your growth channels and shows what high tempo testing can do for you: