Metric Trees: How Top Data Teams Impact Growth

‍

You can have the perfect data stack and still be lacking impact as a data team. Ask yourself, do you understand your company’s growth model? Understanding how your key metrics drive your business forwards is essential to becoming an indispensable data team. On the simplest level, we need to know which metrics are driving growth (our KPIs) but the best data teams also understand how they influence each other, and the component metrics that drive their KPIs. So why aren’t we talking about growth models more in the data community?

‍

The problem is there always exists a gap between what we can achieve with our data model (the tables, joins and aggregations) and what we need to measure and track our growth. The reality of businesses can be fuzzy and hard to reconcile (I’m looking at you marketing attribution). But when our metrics are isolated to teams or specific data models, we can miss opportunities for seeing cross-business changes that often lead to the biggest improvements. For example, we might believe that reducing our response time on customer support would improve product retention. But how does something so fuzzy map into our logical world of data?

‍

Introducing Metric Trees

‍

Today we’re launching Metric Trees in Lightdash, a new way to model your business without the constraints of your data model. Metric trees allow you to map out the causal relationships between metrics without needing to model that. Designing these metric trees on a canvas is a valuable task in itself, and provides an important documentation.

‍

Some of the best data teams are already creating metric trees in tools like Figma and Miro. But a model isn’t enough, you also need to measure growth and with existing whiteboarding tools you can’t answer “how is growth today?” without having to jump to your favourite BI tool.

‍

With Lightdash, you can now see the current state of your businesses right on top of your growth model. Each node on your Lightdash metric tree shows it’s current value and % change over time. The metric tree in Lightdash has been designed to be consumed by anybody in the business and it’s a great jumping off point for a deeper dive into a metric.

‍

‍

How to Setup a Metric Tree

‍

To get started in Lightdash, go to Metrics in the navigation bar to see a list of all of your metrics. Choose a metric category (or add some if you haven’t already) to help you narrow down the size of your tree. A good starting point would be to tag your KPIs and effort metrics. From there click the tree button in the top-right to enter the canvas view (see video at the top).

‍

If it’s your first time on the canvas, you’ll notice all your metrics are waiting to be connected. You can drag those onto the canvas and start connecting them up. This exercise is so valuable and helps align the data teams with stakeholders, so don’t do this alone! Grab some of your key decision makers in the business and go through this exercise together.

‍

Once you’re happy with the tree, share it with your team and they’ll be able to see at a glance the current state of the business and the relationships between them. Remember, your growth model will evolve and so set a reminder to review your metric tree in a month or quarter as the business moves forward.

‍

Use Cases for Metric Trees

‍

Your first metric tree is guaranteed to find some interesting patterns that you wouldn’t expect such as “revenue is up but our churn is higher than expected”. Use that to dig in to the metrics and understand what’s driving the KPIs. Lightdash shows how your metrics change through your dashboards, alerts, and scheduled deliveries but when you want to know why the metrics changed, it’s time to revisit your metric tree and start debugging what the root cause of the change could be. That’s the power of the growth model!

‍