Comment & Opinion

What is a data warehouse and why do hit studios use them?

GameAnalytics' Sara Rubio explains

What is a data warehouse and why do hit studios use them?

Sara Rubio is a data scientist - team lead at GameAnalytics.

The other day, I had to venture into the tangled mess of my cable bag.

No matter how organized I think I’m being – neatly wrapping cords around their plugs and keeping everything separate – whenever I leave the bag alone for five minutes, it all turns into a Gordian knot by the time I return.

It’s infuriating. I really should get a nice box, and make it easy to find what I need.

It struck me how similar that feeling is to juggling data from a bunch of different games. Lots of tools and data, all in a mess.

But thankfully, you can get a data warehouse to fix the problem.

Data warehouses make it much easier to search through and find what you need, reveal insights, and help you combine all those different sources.

They’re an essential part of any hit studio: without a data warehouse, finding the right data is a chore that can waste hours.

So let’s look at exactly how data warehouses work and how to use them.

What is a data warehouse?

A data warehouse allows you to analyze huge amounts of data and create reports. You can bring in and store the data from multiple different tools.

Data warehouses make it much easier to search through and find what you need, reveal insights, and help you combine all those different sources.

Once they’re combined, you can find out so much more than when they’re separate.
In fact, a good data warehouse will let you:

Discover trends in your portfolio

When you bring the data in from all your games, you can find out how players are behaving across the board.

Do you have a problem with day-one retention? Or is it just the one game? How much do you tend to earn from downloads? Does it vary by genre?

Feed into your other tools as a single source of truth

Maybe you’ve got machine learning models that you want to train or data visualization tools you’d prefer to use.

A good data warehouse lets you easily plug into those without needing to develop your own SDKs, ETL pipelines or storage platforms.

Drill down to player-level data

The best analysis comes from deep, specific events, not broad strokes.

vBut capturing that data and comparing it is difficult unless your data warehouse links up your players (or their devices) correctly.

Our Player Warehouse does it all

As part of our Data Services, we’ve created the Player Warehouse: a data warehouse, specifically designed for game developers. We know the kinds of data you want to search and how that data is typically captured, so we know how to convert it into something more usable.

How leading studios and publishers use them

We obviously can’t speak for all data warehouses, but with Player Warehouse you can start running queries from minute one, on your first day.

With so much choice, people keep asking us what kinds of searches they can do (and what the best of the best focus on, too). So here are a few examples of how you might use the Player Warehouse.

  • 1. Target your highest earners

Figuring out the return on your investment for a single device is key to finding your whales. And how they behave. If you know that, you can either focus your campaigns on those players, funnel them to your next game to drive up the revenue, or increase the number of ads you show them.

In short, having this information means you can make more money from your best-performing players.

How it works

Player Warehouse combines the ad revenue and in-app-purchase revenue with your cost-per-click per user. Because it’s all device-level data, it means you can get the return on investment for each specific device ID.

  • 2. Fix bugs faster

When you have all your data in one place, you can search for anomalies. If you can spot these problems before your players start complaining, you can isolate what’s going on and fix it. By being proactive, you’ll lower your churn rate. Lower churn, better retention, higher revenue.

How it works

You can easily run a query to find out which devices are sending the most errors. Once you know that, you can search to see where players get stuck on a level. Now you know where they’re stuck, you can look for that bug and squash it.

  • 3. Teach a machine to spot opportunities

You can feed the data from Player Warehouse into your own machine learning and AI models. The opportunities here are endless. Maybe you’ll teach the machine to help your advertising, building up user profiles. Maybe it helps you optimize your advertising bidding. Or maybe you use it to increase your margins.

You can feed the data from Player Warehouse into your own machine learning and AI models. The opportunities here are endless.

The choice is yours.

  • 4. Spot trends and insights from your data

Most popular visualization tools let you use BigQuery, so you can directly connect them to Player Warehouse and start building your own dashboards. Just focus on what’s important to you, and add as much personalization as you need. If you’re looking for a tool, we’d recommend Tableau, Google Studio, Looker or Holisitcs.

Untangle your data

Our Player Warehouse is an essential piece of kit if you’re looking to manage a portfolio of games and all the data that comes with it.

You can divide up your audience, create more in-depth A/B tests across your portfolio or get more insight into how your players behave.

We’ll handle the data, you just focus on making games that players will love.

Get in touch to learn more about the pricing.


PocketGamer.biz regularly posts content from a variety of guest writers across the games industry. These encompass a wide range of topics and people from different backgrounds and diversities, sharing their opinion on the hottest trending topics, undiscovered gems and what the future of the business holds.