Deriving revenue from in game adverts in on-line mobile games

Highlights

• Revenue from in game advertising is enhanced if adverts are interesting to players.

• To optimise revenue the player base should be segmented by region and device used.

• Successful segmentation can be achieved with public cultural data.

• Segmentation should be validated using multi-level regression approaches.

Abstract

With the increased ownership of smart phones online free to play gaming has become popular but deriving revenue for the game owner is difficult and relies on encouraging micro payments to buy in game tools or cross selling of other products. There is another revenue generating option, which is to show in-game advertising in which the game owner obtains income from an advertiser or owner of the advert if a player views the advert. Typically, a third-party act as an intermediary to provide advertisements from advert selling auction sites to the game. This provider must be able to embed adverts in the game in real time, so must use several auction sites to ensure advert supply. Some of the auction sites are of more value to the game than others and this value can depend on where in the world the game is played, and the device used. To optimise the supply and order of adverts requires geographical segmentation of the players. In this paper, a method to undertake this segmentation using publicly available data is presented.

Introduction

The universality of smart phones is a phenomenon that has propelled the world of computer entertainment especially the free to play online gaming [35]. Gaudiosi [11] reported that revenue from mobile games are set to overtake revenue from console games and reports that Newzoo, a video game research firm, has raised its 2014 global revenue forecast from $21.7 billion to $25 billion, this new forecast being attributed to a 43% increase in mobile games. According to Lofgren [24] typically those who play mobile games were estimated to spend on average $6 per month. Deriving value arises through multiple channels and Marchand and Hennig-Thurau [27] document this process. Sharma and Morales-Arroyo [33] and Roma and Ragalia [32] point out that there are three main routes to revenue; (1) to convert players paying for free to register for premium services at a cost, (2) to encourage players to make micro purchases, such as cheats, weapons, move up a level or additional lives for small amounts of money and (3) to derive revenue from advertising. This latter route is growing in importance and according to eMarketer (2015) has a potential for establishing a $100 billion market. Typically, adverts are obtained from on-line real time auction sites operated by advertising companies the adverts procured are then inserted into a game while a player plays the game. The game designer has control of when and where an advert is shown, and demand is usually triggered when a player reaches a particular point in the game. The standard mode of operation is that a third party obtains adverts and supplies these to a game provider and the provision is immediate in response to real time demand arising as a player plays the game. The advert is frequently a banner type, which, proves information and might open a video. Each time a player clicks on an advert revenue is generated and paid by the advertising company to the game provider who then pays a small proportion to the third-party provider. There may be a threshold when payments are only made above a certain number of clicks.

From the perspective of the third-party supplier, they require to ensure advertising requests are fulfilled in near instant time and revenue is maximised. However, different advertising networks perform differently in different countries and revenues might depend on the game, demographics of players and the device platform (Android or IOS). In this situation decision-making is difficult as demographic data on the players is generally unavailable and the only available data is number of impressions, expected costs per mille (ECPM), fill rates, country in which play takes place and platform type. Hence, the decision is to decide which advertising networks should be chosen for each country and platform. Because of the evolving nature of the market and scarcity of data from some countries, the desire is to segment the countries into homogeneous groups. By using representative groups then, a waterfall of preferred advertising networks ranked by likely revenue generation can be created for each country and platform grouping. In this the advertising network which has most revenue potential is ordered first, and the advertising network is approached but if no advert is available, then the next potential most lucrative advertising network is contacted and so on until the demand is fulfilled.

In this paper, we forward a strategy to segment countries into similar groups. One can use data from market research companies, but this is often not at a suitable macro level to allow emerging countries to be compared and very specific market data is either not available or is unreliable for countries in emerging markets. We investigate how solutions can be developed from publically available macro data. To present this case we first give a short literature review followed by describing our method and data used. Next, the method exemplified by applying it to two games as an example and the paper closes with a discussion of the benefits and limitations of this approach.