Data can be quite overwhelming and messy by nature, so starting with small process changes is the best way to start. Once you have an easy way to implement tracking, adopting the three following patterns will help your team get into the habit of discussion and experimentation. Here's three ways of thinking about data and UX flows that have significantly sped up and simplified our feedback loop @ Contra. Hi there!

While some UX flows can grow into 10+ steps, its great to be able to boil them down into general, large-step funnels so the team can have a sense of conversion for the full flow and only dive in when necessary. For example, let's say you have this funnel leading from sign up to the key action of creating a blog post:

While the above might be how you initially think about a feature or a series of steps, you should go higher up. Once you encounter a drop-off right at a general spot, you can start getting granular.

The extra benefit is you will start naming sections of flows and this will also solidify understanding of sections of the product that will be very handy for more than analytics.

A super easy quick win is to define the key actions that directly relate to your north star ⭐️, then go into your analytics tools and create two things:

A custom event that groups all your key actions (Example for an e-store: save item to cart, checkout, share item)

A cohort that uses the prior custom event to segment out your power users

Your power users probably take the shortest paths around your app and you need to know see why. Being able to look at them against the rest of the user-base will help you zone in on the behaviors where you want to reduce friction.

Entry points are everything! But often they are not easy to find. If you are using Segment or a similar analytics tool, you'll find that UTM parameters only make it into the initial events on the page where the user lands. This makes it super hard to pin acquisition easily in reporting tools like Mixpanel/Posthog for entire sessions. Make things easier for your team with a really small engineering lift:

When your app loads, save any UTMs that are included in the URL. The most important ones are "utm_source", "utm_medium", "utm_campaign".

Analytics tools use a context property to store and manage UTM campaign data. This context takes has a "campaign" property that in turn should have a source, medium, and name fields.

On every track/page call, you should populate the context property with the data you save.

Voila! Now you will be able to tell where a user came from no matter what point in the session you are curious about.