PlayerZero was designed from the beginning to streamline some of the most fundamental decisions product & engineering teams have to make every day about their product quality. A key feature of PlayerZero’s design is around the capture and inference of identity to build contextual relationships between issues and the things that they may affect.

At it’s core PlayerZero is a correlation engine between two types of streaming events:

  • Product analytics data
  • Engineering monitoring data

With PlayerZero fully configured, the end result is a system that accounts for the errors and other product quality quality signals in your application and semantically tells you the impact of those signals in terms of their effect on analytics events. For example, if you had a Javascript error that was happening on the profile page, PlayerZero would be able to detect that users who run into that Javascript error often can’t continue to save their profiles a few steps down.

This is useful because it allows us to surface the issues that are causing a measurable change in behavior. We call these high impact issues. And just as this mechanism allows us to positively select issues that are impactful to your users, the same mechanism also allows us to negatively select out issues that don’t seem to bother your users too. Together, these properties enable PlayerZero to automatically curate & notify you when new, impactful, and interesting issues are happening…and then can tell you why they’re interesting as well.

Building Blocks

In order to achieve this, PlayerZero has to build an intuition about what the user should have done if the error weren’t present, and then compare that to what they actually do.

This implies a few architectural pieces that are central to understanding & building PlayerZero.

Identity

As we mentioned earlier, identity is a core part of what makes PlayerZero special. Wether you propagate your own application identity with playerzero.identify() or you use our default anonymous identity, PlayerZero in the background is tracking the story of that identity over time, and comparing it live to the behavior of all of their peers. This means that if we take a signal processing view of PlayerZero, we’re operating over identity to achieve our result as opposed to other monitoring tools which operate over the error itself as their core unit. This piece of architecture, while subtle, actually underlines the entire philosophy at PlayerZero.

A question we often find ourselves asking is:

If an issue doesn’t impact anybody, is it still an important issue?

It’s a philosophical question at heart because the terms “impact” and “important” are truly subjective, and their definitions live in a select few people’s mental space. Chances are if you’re reading this, you’re probably one of those few people.

But if you’re trying to bring the entire team on the same page (literally), we have to be able to tell a compelling story that resonates with anyone.

Identity (even anonymous identity) is important because it enables PlayerZero to take an opinionated stance on this topic.

We believe that issues become important when they begin influencing a measurable change in your system’s behavior. With analytics, we can measure that change, and with identity we can tell the story of that change.