Big Data applications for game marketing

I recently conducted a survey where 1,400 gamers answered questions about their game buying and playing habits. In the study I was particularly interested in asking about the Pile of Shame: a gamer’s backlog of purchased yet unplayed games. The study yielded many interesting results and you can read about how the current age of the Perpetual Sale affects gamers over on Kotaku. An infographic with highlights from the study appears at the bottom of this post.

For professional game developers, and especially those responsible for pricing and marketing games, it yielded some interesting nuggets of insight. The average gamer surveyed buys just 20% of games new at full price vs 60% on sale. Of the 1,400 gamers surveyed only 31% buy at least half of their games new at full price. As I explain in the Kotaku piece, gamers have been trained by Steam Sales, Humble Bundles, PS+, used games et al. that games will be cheap if you are willing to wait a few weeks or months. As one survey respondent wrote “I will only buy video games on sale.  All games go on sale, given time.”

More gamers are spending more money on more different types of games, which can only be a good thing in my opinion. As both a gamer and a developer I appreciate the incredible range of gaming experiences coming from developers of every shape and size. The problem for some is that gamers are spending that money in smaller and smaller increments.

For developers at publicly traded companies these stats are likely to induce panic. In a world where this quarter’s earnings report is the difference between bonanza bonuses and studio closures, new game sales are all that matter. Wall Street does not have the patience for long tail successes produced when the majority of a game’s revenue can come from sales and bundles months or years after initial release.

Most of the analysis I performed on the dataset produced intuitive results (for example, gamers who buy a high percentage of games on sale are more likely to not play games purchased). But there was one interesting correlation that is of use to anyone whose job depends on selling games at full price.

In the data, I found a moderate correlation (r = .38) between gamers who purchase games at full price and gamers who complete the games they purchase. The type of gamer who plays games through to the end is more likely to buy games new at full price. For those working on franchises, especially annualized ones, this could be a key to an effective marketing campaign.

Much is made of the importance of Big Data, but from my personal experience many game developers are at a loss when it comes to using the metrics they may be collecting in their games to improve the game experience. Many are dubious of the power of data to produce insight. “Game development is an art not a science” is their battle cry. It is even more unlikely that the marketing team responsible for selling the game has ready access to game based telemetry, or can tie that data to email addresses that can legally be targeted with a special campaign or promotion.

But in a world where a marketer had such perfect insight, she could develop a special campaign targeted at those gamers most likely to buy on day one. Perhaps all gamers who finished the last game and/or activated an account within the first month of release are invited to a special forum where they can interact with the game developers, get early access to beta tests, earn unique items for the game and compete for the chance to have their name or other personal mark in the game. Instead of being sent from a generic corporate account, this invitation could be send by and in the voice of a well-known figure on the game development team.  

For example, I have purchased new and completed every console Ratchet and Clank game since Up Your Arsenal. I love Ratchet and would readily jump at the chance to help shape a future game in the series as a fan. If Sony/Insomniac could tie my trophy data to an email address and reach out to me and other Ratchet fans, they could craft a campaign specifically targeted to energize the base and “get out the vote” on day one. I would love to get such an invitation, especially if it was sent from Brian Allgeier and not just Insomniac Games.  Instead, I almost missed the launch of Into the Nexus in the avalanche of end of year releases.

For indie game developers struggling with the tricky issue of pricing a game, the study can provide valuable context to help make the pricing decision. When the average gamer buys 60% of his games on sale, a developer needs to price a game with discounting in mind and set revenue expectations accordingly. Not all developers are lucky enough to get featured in a Steam daily or holiday sale, but it is clearly the goal state for many devs. When pricing your game, the question “How will it look at 75% off?” should be a careful consideration.

Although my study into game buying habits produced some interesting results, I am left with more questions than answers. I have posted a new survey about the habits and motivations behind purchasing and playing games, and look forward digging into the data and sharing my findings with my fellow gamers and developers.

A note on research methodology. The survey was conducted online using Typeform and publicized largely through Twitter and Reddit. No demographic or console preference data was collected, but the dataset likely skews towards PC gamers in North America and other English speaking countries. All averages reported in the article and infographic are median values and not mathematical averages; a small number of outliers with massive gaming backlogs would otherwise skew the data in non-representative ways. r values alluded to are Pearson's correlation coefficients calculated in Excel.