Intro to growth engineering - learn how to prioritize the right projects
About the author
Jeff Chang (@JeffChang30) is a growth technical leader at Pinterest and angel investor. If your startup is looking for an angel investor who can help with all things growth, please send over an email!
A good growth opportunity is an area of optimization that can lead to a sustainable increase in growth rate. To make a product grow, usually one area of growth is not enough, and having many areas diversifies your bets to reduce the impact that losing one area will have. Popular growth opportunities include SEO, signup/login conversions, performance, and mobile acquisition. In this series, I will assume you have a way to run AB experiments and are just starting to ramp your growth engineering. This post will talk about how to pick a growth opportunity to work on and how to get started, while future posts will deep dive into specific ones.
Which growth opportunity should I choose work on?
The answer to this question will come from your data. First, you want to pick some sort of north star metric that you want to grow. Note that this metric can and probably will change over time, but optimizing for this metric now will usually have positive directionality. Common north star metrics are weekly active users or subscriptions. Here is a simplified user state diagram:
You want to increase flow into your core action, and reduce user dropoffs and dormancy. To measure acquisition opportunity sizes, you need to know your unauth traffic, signup conversion rate, and activation rate. To measure retention opportunity sizes, you need to know your number of dormant users and resurrection rate. Then, you can just do basic math to estimate the opportunity size. Here is an example of measuring the opportunity size of various opportunities:
Traffic: 10K sessions / day
Signup conversion rate: 5%
New core users / day = 10K sessions / day * 5% conversion rate = 500 core users / day
SEO: Fairly confident we can improve traffic by 3K/day in 3 months, which equates to 150 core users / day
Signup conversion: Fairly confident we can improve signups by 20% in 3 months, which equates to 100 core users / day
Paid acquisition: Fairly confident we can acquire an additional 1K traffic/day in 3 months, which equates to 50 core users / day
Based on these calculations, we should explore SEO traffic opportunities first, then signup conversion. Of course, the most difficult part is coming up with accurate predictions of impact and time required, and that just comes with having more experiment results, understanding the space more, and talking to other growth teams.
A common problem that new growth teams have is that they over-invest in a growth opportunity that is not paying off, while there are many other good opportunities lying around. What I would recommend is if the growth opportunity is not paying off anywhere near expectations, fail fast and try some other ones. Naturally, as the growth team grows, it will invest more into opportunities that continue to show high impact. In the beginning of Pinterest growth, this actually happened quite a few times where we had decent sized teams working on a project for months, with little to no impact. Sometimes, a project would be completely scrapped and months of work would be wasted. Since then, we’ve learned to fail fast and now every subteam on growth has a significant amount of impact.
As for specific opportunities, I would recommend trying at least one acquisition opportunity and one retention opportunity to start out. Acquisition without retention has a well known “collecting water into a leaky bucket” problem, and retention without acquisition leads to slow growth rates. So, you will eventually need to solve both problems. Here are some sample opportunities in each category and some product characteristics that suggest you should try each one.
SEO - Product has quality web content
Mobile web to app Conversion - Product gets mobile web traffic and also has an app
Signup Conversion - Every product that requires authentication
Performance - Every product
Paid acquisition - Company has a budget and decent LTV
Virality - Product has social aspect
Referrals - Product has something to give to users to incentivize them
SEO - Product has a lot of web content
Email - Every product
Performance - Every product
Push notifications - Product has an app
Login conversion - Every product
Paid resurrections - Company has a budget and decent LTV
As you can see, a lot of these opportunities probably apply to your product, and this list is not exhaustive. Eventually, you want to have tried them all!
I picked a growth opportunity to work on, what should I do next?
"The biggest pitfall I see when people work on new opportunities is too much planning and not enough experimenting and learning."
Planning is fun. It’s fun to brainstorm and come up with a bunch of cool ideas. It’s fun to set a goal and try to make a plan in order to reach that goal. However, time spent planning is time spent not learning. The plan you initially come up with probably isn’t that good anyways, given it’s new. In fact, even now when asked what my team going to work on 6 months from now, I usually say that we have options but don’t have a set plan. This is because we will be smarter at picking which opportunities to pursue 6 months from now than we are today. Even though we’ve been working on growth for many years, we still continue to learn a lot, so having a learning focused mentality is especially important for newer growth teams.
So, the first thing you should do when working on a new opportunity is simply to run some quick experiments. A lot of the time, it doesn’t take some hard insight to be able to rack up some easy wins. For example, if you are exploring signup conversion, it’s very easy to run a lot of basic experiments as long as you have an AB experiment framework. A lot of these experiments will fail, but what’s important is that this information is vital to when you actually start creating your plan. To come up with experiment ideas, I have three main sources of inspiration: see what other products do, look into the metrics to find opportunities to improve, and go through the flow as a user. Once you have a couple of experiment results, now you can assess whether or not the opportunity is worth pursuing. If you have some wins, now you can plan with much more insight into what works and what doesn’t.
Growing growth opportunities is a hard skill to learn, but very valuable because it can completely change the trajectory of a company. Hopefully from this post you learned how to pick a good opportunity and get started! In upcoming posts, I will talk more about various topics to help you scale your growth with great engineering and process, and also talk about specific growth opportunities. The first one will be the popular SEO channel. Stay tuned!
Compare opportunities using data to determine which to pursue first
Experiment and learn instead of creating a master plan in the beginning
Acknowledgements: Thanks Casey for great feedback on this blog post!
Want some advice about getting started with growth engineering? Email me at firstname.lastname@example.org
Scaling new growth opportunities series