Skip to main content

Alan Miller: Part 1

The Activision co-founder on why he's joined the GamesAnalytics team

With the emergence of online - and increasingly with the importance of social network games - understanding your game's users has become a hugely significant element of the business, with several companies now offering analytics services to developers and publishers.

One such company is GamesAnalytics, first announced last week, but today they bolster their team with the appointment of Activision co-founder Alan Miller to the role of strategic advisor and director of North American operations. Here he explains why he's getting involved, and precisely how analytics can help. The press release announcing GamesAnalytics last week talked of "turnkey real-time behavioural analytics" - let's start by unpacking what that is.
Alan Miller

Sure, I'd be happy to address that - perhaps we should break down the phrase. "Turnkey" means that we provide a solution that is extremely easy to integrate; "real-time" means that we're constantly analysing player data and providing individualised marketing to players through in-game mechanisms.

The objective of all of this is to better understand player behaviours, in order to help publishers maximise revenue; and from a game-player's perspective, to minimise frustration and increase the length of time they're interested in playing the games. And what kinds of things are you tracking? When you talk about behaviour, what are you looking at?
Alan Miller

Well, we use very sophisticated analysis to look at the entire player data-set, which could be hundreds of gigabytes or terrabytes in size. We look at it in an abstract way to identify desired behaviours.

For example, the common problem in the industry is that only 1-2 per cent of online players generally make payments - and a lot of publishers are resigned to that. But we look at that as a great opportunity to bring in additional paying players.

1 per cent of those 98-99 per cent of non-paying players substantially increases revenue, so if we assign our analytics to look for behaviours that address additional monetisation, we can identify several different patterns of behaviours that players have experienced in the past that lead to this monetisation.

Then, in real-time, we look at all the active players, see if they're close to this monetisation model - and if they're close to it then we give them very specific marketing messages, like hints, challenges, offers, free goods, instructions and so on, that are delivered in-game to the player to motivate them to move towards these behaviours.

The desired behaviours don't necessarily have to be monetisation-related - the objective might be to study the sending of messages to other players in order to increase virality, for example. We're working with some clients right now to help avoid players from leaving the game - cancelling their subscription and departing.

So there are several different kinds of these desired or avoided behaviours that we study - and maximise the results for the publisher and players.

Related topics