Data Insights in Live Ops: Track Less, Understand More

It is tempting to track every metric a game can produce. More data feels like more rigor. But Douglas Adams settled this decades ago: a supercomputer spends millions of years calculating that the answer to life, the universe, and everything is 42, only for everyone to realize they never worked out the question. As a certain supercomputer put it, only when you know the question will you understand what the answer means. Data works the same way. A number you cannot tie to a decision is not insight. It is trivia.

Data Insights, the fifth of the 7 Pillars of Live Ops, is not about collecting more. It is about understanding more, and making sure the whole team understands the same things. That takes three moves: tracking with intent, combining the right kinds of data, and putting the result in front of everyone who makes decisions.

Start with the question, not the metric

The instinct to track everything comes from a good place, but it produces dashboards nobody reads and a false sense of being data-driven. Data is only powerful when it is actionable, and actionable starts with a question, not a tag.

So before you instrument anything, ask two things. What decision am I trying to inform? And what player behavior actually moves that decision? A handful of metrics chosen that way will teach you more than a thousand collected on reflex. Focus produces depth: you understand what you measure, you prioritize better, and your live ops get sharper instead of drowning in noise. The studios that track less, deliberately, almost always understand more.

Tracking everything is a privilege, and a liability

There is a version of this where you do track everything and point AI at the pile to surface patterns. It works, but be honest about who can actually do it. Only the largest companies have the financial and technical capacity to capture and mine every action a player takes. For everyone else, exhaustive tracking is a cost you cannot justify and a distraction from the questions that matter.

And even the giants pay for it. In 2025 the privacy NGO noyb filed a GDPR complaint against Ubisoft over Far Cry Primal, a single-player game that still required an online login. A player's traffic analysis reportedly found around 150 DNS requests and 56 connection attempts to external servers in just ten minutes of solo play, with data reaching third parties such as Google, Amazon, and Datadog. Ubisoft's position is that the connection verifies game ownership; noyb argues that Steam already handles that, so the collection is not necessary and therefore not lawful under the GDPR. The case is still open, and it is worth following, because it draws a hard line: the more you collect, the more you have to justify, secure, and defend. Focused tracking is not only cheaper and clearer, it carries less risk. This is where data meets ethics, and treating your players' data with restraint is part of doing live ops well.

The 360 degree picture: three kinds of data

Tracking with intent tells you what to measure. The next question is what kinds of data you actually need, and the answer is never just one. A single source lies by omission. A real picture, the 360 view, combines three.

Quantitative data is what players do: the behavior, the funnels, the retention curves, the actions logged in the game. It is the ground truth of what is happening.

Qualitative data is what players say and feel: reviews, surveys, support tickets, community sentiment. It is where the why lives, the reason behind the behavior. But as covered in the Live Development pillar, qualitative input is only as trustworthy as your ability to attribute it. A voice you cannot tie to a known segment is an opinion of unknown weight, which is exactly why surveys and VIP programs matter: they let you hear from the players who actually drive your numbers.

Business KPIs are the honesty check: how much players need to spend, and how long they need to stay, for the game to remain profitable. Without this layer you can optimize engagement straight into a game that everyone loves and nobody can afford to keep running.

None of the three is enough alone. Quantitative without qualitative tells you what happened but not why. Qualitative without quantitative is anecdote. Either one without business KPIs is activity without sustainability. Combine all three and you finally understand the why behind a decision, which is the entire point.

A number means nothing without a target

There is a second job KPIs do, beyond keeping you honest about profitability: they are your benchmark. They are how a team agrees on what good actually looks like.

Raw data, however polished the dashboard, cannot tell you whether you are winning. Day-one retention of 10 percent is just a number. Triumph or disaster? You cannot say until you have set the target. Against a goal of 25 percent it is a win worth celebrating; against 45 percent it is an alarm. The KPI is the line that turns a measurement into a verdict.

This is why data without goals quietly fails a team. Everyone can stare at the same chart and walk away with a different conclusion, because each person is silently scoring it against their own private expectation of what good means. Set the KPI, agree the target, and the chart suddenly says one thing to everyone. The goal is the shared yardstick, and it is what lets a whole team look at the same data and genuinely understand the same thing.

Everyone, not just the analysts

All of this only becomes a culture when the picture reaches everyone. A 360 view and a clear set of goals that live only in an analyst's spreadsheet change nothing. The whole team has to have them, readily, so they understand the why behind each call.

That is what actually ends decision-making by gut feeling or by the loudest voices: not a better argument, but a shared, visible picture everyone trusts. It is also how you build genuine conviction in a plan, because a team that can see the evidence owns the decision instead of merely complying with it. In practice this is concrete work: the tag plans that decide what gets measured, the dashboards (a Tools job) that make it visible to everyone, and the 360 reports that turn raw data into a story the whole team can act on.

Where this fits

Data Insights sits in the operational backbone of the framework, the pillars that power everything players actually see. It is what Live Development prioritizes from, how you know whether a monetization model is working or just extracting, and how you confirm whether players genuinely feel heard, which is what makes Communication honest rather than performative. It runs on the Tools that surface it, and it only changes anything when the whole team can see it.

A live ops audit looks at whether you are tracking with intent, whether you have all three kinds of data and not just the easy one, and whether that picture actually reaches the people making decisions. Often the fastest win is not collecting more. It is connecting what you already have.

How do you decide what to track, and has cutting back on data ever made your decisions better?