Some research research from Microsoft has suggested that free ad-supported apps can drain a smartphone’s battery considerably quicker than the paid ad-free counterpart. There are few surprises here, as the free app would need extra resources to download the ad graphics, serve and track the ads, and possibly leave the 3G or WiFi connection open for a few extra seconds after serving the ad to offer a quicker connection to the advertiser’s website if the ad is clicked. However, the scale of the extra battery drain is something worth exploring.
The research [PDF], authored by Abhinav Pathak and produced by a team at Purdue University in Indiana, USA, suggests that up to 80% of the energy consumption of Angry Birds free was from supporting advertisements and user-tracking. Such an increase is startling, and a good guide to app developers to make sure they keep an eye on these 3rd party tools when integrating them into their apps. This large an increase can only really be explained by user and location tracking within the app, as it is more important to track how users actually use an app when it is ad-supported to be able to sell that information to advertisers, whilst a paid version is ad-free and so such information is not required.
It was the geolocation tracking via GPS that the researchers blamed for the major increase in battery usage, with the location checked and sent to the ad server each time the app was opened, or even during usage. They do not say which mobile ad service performed best or worst in these tests, but one would imagine that the location of the user would only really need to be checked once every month or two by the app – as ads targeted to their home location are probably more relevant than those targeted to them in a week on holiday (when data roaming may well be turned off anyway). Whether this battery drain is universal after first use does not appear to be investigated, as one would imagine such details would be cached locally.
The study was limited to investigating Android and Windows Phone 7 smartphones, as the researches could not gain the same access to iOS powered iPhones, but in all likelihood the findings would be similar across the board with the possible exception of apps using the built in iAds in iOS which may well be better optimised by Apple itself.
The granular level of investigation here is fascinating and very useful data for app developers to make use of – but we need further investigation with a focus on 3rd party ad integration systems to really move forward here.