Growth Function – according to YCombinator

I’ve been doing a lot of reading on what a Growth team should be doing and here is my ramblings and thoughts based on several hours of Podcasts from YC:

Set and Optimize Key Growth Metrics along the Funnel

Borrowing from Dave McClure’s Pirate metrics, one of the first things I did is to think about our customer journey as a funnel and to measure different dimensions of the funnel:

Within my B2B context, I imagine this would look like this:

  • The first step in this journey is Acquisition, which in a B2B context, involves installation of our product onto a user’s infrastructure and creating a user account.
  • The second would be Activation, which in a B2B context involves sufficient configuration and setup of our product to the point where target users understand what our product does and have an “Aha!” moment.
  • The third would be Referral, where in a B2B context the user would tell others (most likely their boss, or their purchasing team) about how great our product is. There are other parts of the funnel (Retention and Revenue) but I’m going to focus on the first aspects which I’ve characterized as Growth.

With the journey framed in this manner, as opposed to with traditional SaaS metrics like ARR, Churn, MRR, and Pipeline, I’m able to build out a strategy and roadmap of products, features, and enhancements that make each of these areas more efficient and have less friction.

Experimentation, aka Many Tiny Failures

Dave Brueck introduced me to the concept of Many Tiny Releases back in 2015, which I now understand to be also known as CI/CD and is followed by almost all modern software engineering organizations. After listening to Y-Combinator’s Growth office hours Part 1, 2, and reading their Growth guide I feel that the second goal of the Growth team is essentially to try lots of different things and learn through success and failure in the field. Borrowing from YC:


What I’m not sure about is how much of this is feasible / manageable at a B2B SaaS company where the user base is typically much smaller than the consumer properties where Growth teams are most prevalent (i.e. Facebook, AirBNB, Uber, etc). The number of MAUs that would be at a B2B SaaS company would typically be orders of magnitudes smaller then a consumer facing property (with some exceptions such as Slack). On the one hand, it will be much easier to actually run and keep track of experiments since the volume of data will be more manageable. That said, the feedback loop is going to be longer and the statistical significance of specific experiments is going to much less due to lower volumes.

I suspect there’s some things I can try, but at this point I don’t know the best tools or processes to manage this.

Tooling, Communication and Transparency

In addition to having tools that can measure various results of experiments, I believe that it is equally important to set up dashboards, communications, and protocols that can be used to communicate the results of the experiments back to the team and to stakeholders. This will help drive focus and accountability back to the team, while at the same time raising the profile of the various things the team is trying so that if any of them are radically successful, the organization can rally around them.

YC Growth Guide

For a company operating at the scale that we are, this is going to take investment. Building and hardening the appropriate collection, analysis, storage, and streaming infrastructure is a significant effort that is going to an engineer or data scientist to get the plumbing and systems set up, and a data scientist to create the first few views. Building out the right team here is critical for future flexibility in goal setting and metrics.