“The question I always get is, ‘What is a numbers and statistical learning nerd doing in marketing?'” says Kiyoto Tamura, vice president of marketing at Treasure Data. “I feel like I’m on a personal journey to close the gap between the hype and the truly important underpinnings of the current AI revolution, and that is data.”
Originally an algorithmic trader in Chicago, using algorithms and data processing to make money, and then a backend anti-fraud engineer at TrialPay, Tamura found that the marketing world was not embracing the potential of the kind of powerful algorithms hat he was creating and learning from the most comprehensive, up-to-date data possible.
At Treasure Data, Tamura focuses on taking games like Survios’s Raw Data, the first VR-exclusive game to reach number one on Steam’s global top sellers list, to the top. And it all comes down to data.
“We’ve always taken this belief that we go wherever interesting, relevant, and new data is being generated,” Tamura explains. “What we find interesting about VR and AR is, as a collection mechanism, it’s still going to be use the same networks and the same core technology, but the type of data you’ll be able to collect will be unlimited and unforeseen.”
However, he says, the industry is still in its infancy, and at this point, the significance of every piece of data collected might not yet be understood — but collecting as much data as possible and creating a repository of raw data is essential even at this stage of the game.
Tamura points to Wish, the mammoth mobile shopping site, and one of their first customers, as an example of how important comprehensive data collection is. From the start, he says,the company instrumented everything, from who is logging in, how they’re logging in, how they’re interacting with different SKUs in different geos and different zip codes and more.
“Fast forward four or five years, all these iterative decisions they’ve made based on a very KPI-driven approach have taken them from a little startup in San Francisco to billions of dollars in valuation,” says Tamura.
But no matter what metrics you decide to focus on, there is one absolute must-have data point VR game companies need to capture, and that’s time spent.
“Some people call it engagement; some people call it retention — but I feel like those are just translations of a core idea: what fraction of your customer’s time is spent on a system that you built, so that data comes to you and becomes your business’s asset?” he explains.
That means figuring out who these people are who continue to come back to a game and invest a consistent chunk of their time — day in, day out, week in, week out — in your game. You’ve unlocked something about how they like to spend their day.
“Every time we use a device, we generate data,” he adds. “That means the more time of someone’s data you have as a product company, the more data you’re going to accrue, and the more data you accrue, the more data you can use to train all these cool AI innovations coming out, and the more sustainable your advantage is going to be. That is really the key.”
There are a growing number of tools to help companies hook into the data sources you need to collect data — including scalable ingestion, scalable storage, scalable processing, and machine learning for the data you store.
“The key factor a company needs to consider when choosing with any solution — whether that’s a marketing analytics solution or a product analytics solution or building a lot of these by yourself and assembling components – is the long-term benefit, which is data ownership,” Tamura says. “Yes, focus on immediate needs and business impact, but always pay attention to, ‘What if we grow out of this system? Can I take my data that I have stored or entrusted to that system with me so that I can continue on my analytics journey without having my data held hostage?”
Security is also essential he adds, and is becoming increasingly important — especially with the GDPR being enforced next May (the General Data Protection Regulation being introduced in the EU). So it’s more important than ever to consider the security impact of how you collect and store customer data.
“Buyers need to pay attention — do those folks have the right security compliances and certifications? And what is the actual practice of using data?” he says. “Even if you use the most secure, most compliant tools, if you don’t use them right, you expose yourself to a lot of risk.”
To learn more about how and what data you need, the KPIs you should be stalking, and how to full-on deploy full-stack analytics infrastructure and tools, don’t miss this VB Live event.