Growth Hackers Should Think like Scientists
I’m a firm believer in that growth hacking is about constantly experimenting and learning from successes and mistakes. In fact, I believe good growth hackers are actually scientists at heart, with a marketing veneer. For example, the growth hacking process that we’ve established at Etsy draws parallels to the Scientific Method that I learned in junior high:
- Ask a Question — how can we bring more buyers back to Etsy everyday?
- Do Background Research—what have we tried in the past that worked? What have our competitors tried?
- Construct a Hypothesis—if we send our buyers personalized and interesting content, they will click through and come back to the site.
- Test Our Hypothesis by Doing an Experiment—build a backlog of features/hacks we want to test; make sure to keep the scope of the test small so we can narrow down the winning/losing ingredient.
- Analyze Our Data and Draw a Conclusion—did we see an impact? Was it significant?
- Communicate Our Results—record our results on a wiki and send out to our peers so they can learn as well.
A good growth blog taught me that building a successful growth team is about making this process quick, repeatable and iterative. I am still refining the process for my team, but here is what we have so far:
- A knowledge repository containing every UX research and product experiment my team has ran in 2014, with where we failed, where we succeeded and what we learned.
- A prioritized experiment backlog. The Head of Growth at LinkedIn once told me that hacking growth is about taking a portfolio based approach: always make some small bets alongside your big, long term investments. I like to use a 2x2 matrix to categorize our experiments:
In this example, the small bets are the “Fill Ins” and the big investments are “Major Projects.” I aim to keep a healthy proportion of both in our backlog.
- An experiment template to record hypothesis, results and analysis in a consistent and readable manner.