Starting A Growth Team: Ten Key Lessons Learned
It was about 12 months ago when I was challenged by our CEO to dismantle our squad and create a growth team. Not being too sure what a growth team was, we started anyway. We got access to work completed by Andrew Chen & Brian Balfour, took a few courses from the Reforge series and started developing our growth loops. It all seemed so straightforward and I was so naïve! The more we got into the weeds of these growth models, the more we found areas of data that we did not have access to, parts of the business that were run in silos and issues with bringing it all together.
The idea was that the growth team itself would be a 6 month experiment to show: 1. we could create enough value, 2. we could get internal buy-in across the organization and 3. we could propagate a culture of data driven experimentation to other teams. By the end of month 5 I could see that we had achieved 1 and 3 but judging the success of 2 was a bit harder. It took many honest conversations at multiple levels of the company to see what was going right and what was not. The reason that getting internal buy-in proves hard for a Growth team is that not only is it a newish concept but it also overlaps with too many well-defined functions. Its scope is large but focused and ever changing.
10 things I have learnt in this journey so far:
- Get everyone aligned to the team’s objective- It was very beneficial for us to have initial conversations with stakeholders to define what value the team would create. We also realized that we were not going to have a clear path from day one and these conversations would have to be revisited regularly.
- It takes a village! Coming up with a large number of hypotheses, across various functions and actually testing them on top of everything else that is going on in an organization takes a lot of communication, collaboration and alignment. The biggest opportunities are often at intersections of various silos where no one is looking. But, by nature, these are a lot harder to tackle.
- Analysis first- Best ideas come from a good understanding of the business and the customers. Not just on the surface KPIs, but cross sections of interesting segments, behaviour and cohorts. It takes both quantitative rigor and qualitative empathy to correctly correlate an anomaly in the data to a hard-to-observe customer pain point.
- Be methodical and diligent- We created and followed processes for ideation, prioritization and analysis to ensure that we are working effectively and spending the required time to fully internalize the learnings from each experiment.
- It takes time to build confidence- With any new function or team, it takes time to have consistently communicated gains before the wider organization believes in the team’s value. Even though the start of the process was the hardest and when we needed the most support, it was helpful to start with focusing on goals that we could tackle with minimal support to prove the team’s viability and encourage others to engage with our initiatives.
- So many ideas, so little time- Once we got used to brainstorming and designing experiments, we wanted to test everything. To ensure that our time and resources were spent efficiently, we had to create a framework to prioritize these ideas based on potential value to the business, effort and their strategic significance
- Success= the right idea + right execution- We have seen time and again that great ideas with weak execution fail without exception. The only way to give an idea a fair chance is to make sure the experiment is designed with execution and data collection in mind.
- Timing matters- As careful as we are with our testing design, we only control a small part of a very large ecosystem that includes other departments’ initiatives, competitor efforts and overall economic conditions in our industry. It is important to recognize other forces that can affect customer behaviour more than our tests.
- Don’t copy from others- It’s great practice to try to understand why our competitors do certain things and we can learn very valuable insights. But blindly copying just doesn’t work. Even if your competitor is in many ways similar to you, there are always differences in how customers perceive your product or what they expect. Instead we keep an eye out for interesting ideas and always test the best industry practices to see whether they work in our context.
- Come up with the hypothesis first- We never start a test without a documented hypothesis. Customer behaviour is hard to predict and there is so much that we haven’t learnt yet, so we don’t assume that our ideas will work. Also, positioning our ideas as hypotheses help to explain to stakeholders the value of experimentation and lowers our reputation risk.
It is an understatement to say this has been an interesting journey with lots of challenges and learnings. The Covid-19 pandemic has changed the environment within which we operate and as our customer base grows rapidly, we have new challenges to tackle and experiments to run. Here is to another enthralling year!
Special thanks to Ignacio Colino for joining me on this journey.
