Getting the Measure of Mobile Advertising

With the booming app economy, strategic mobile marketing is more than ever at the forefront for those wanting the biggest bang for their advertising buck. It is the new frontier, with millions of pounds currently being shot into the black hole of mobile advertising, without the ad analytics to prove they are hitting their target.

Advertisers in the UK market are already demanding transparency and accountability for their mobile campaigns, not only to justify the advertising spend to clients and senior management, but also, to make sure they are actually effective. The challenges faced by the industry are not only about standardizing KPIs, but also, about transparency, and the actual ad verification business model for in-app advertising.

Conflict of interest
There is an inherent conflict of interest in mobile advertising when using a pay-per-click model, if the people telling you how many clicks or impressions you get are the same people you are paying for the advertising space. As with any good business model, third party auditing is needed to verify the numbers. With third party verification, advertisers are able to base their advertising decisions on fact, rather than faith. The advertisers are then in control and own their own data.

There are also some ad networks that sprung up during the start of the mobile advertising boom who created their own form of ‘home brew’, self-made SDKs and ad servers that really don’t support robust verification. It makes it difficult to trust the data if no one else has verified it.

So if different networks define metrics in different ways and provide their own data, how can we truly compare performance and know which targets we are reaching with our mobile ad spend?

Fortunately, the industry is moving forward in the UK on the question of standardised KPIs for mobile advertising, as these hot topics are being addressed, or at least discussed by industry bodies. The IAB’s latest and revised Mobile Web Advertising Measurement Guidelines Nov 2012 goes a long way towards refining definitions and standardising metrics for clicks and impressions… but this is only for mobile web advertising. Currently there are no standards and no guidelines for in-app advertising, although rumour has it that the IAB US has a draft in the pipeline.

Technical hurdles
So why is standardising metrics for in-app advertising so challenging? Clicks, impressions and other in-app mobile advertising metrics are measured and tracked based on SDKs. With mobile web advertising, there are only a handful of browsers, such as Safari or Chrome, where the standards need to be set. When we talk about in-app advertising, it is like trying to set mobile web specifications for 50 different browsers - this is technically complex. The in-app advertising tech space is fragmented, and there are real technical hurdles to standardising what happens in the advertiser SDKs.

There are dozens of different SDKs used by different ad networks, and they don’t all work in the same way. They have differing technical capabilities, and often track slightly different metrics. Most of them are closed source, so external tech people can’t even look at the code and work out what is going on. This makes it is difficult to find a standard tracking solution across all of the networks.

The most common metrics ad networks offer are simple clicks and impressions. However, because of the different classification methods and definitions of what exactly is an impression or click, it makes comparing ad networks on the basis of their self-provided data almost impossible.

In my view, the ideal impression verification needs JavaScript to be executed and run. The ad creative must be there long enough in order to be rendered, so that the average user can see and recognize it.

Yes, there are differing methods for counting clicks. In addition to defining and standardising clicks and impressions, physical limitations (big finger, small screen) and fraud are challenges. Accidental clicks from lack of human dexterity can be removed from the metrics equation by disregarding all clicks that occur within the first second of an ad being shown. As fast as we like to think we are, not many people can see, register and then click on a mobile ad within a second. Verification ensuring a unique click rate is also important, as it cuts out double clicks. Technology has significantly advanced in the past year and advertisers should be able to expect meaningful, verified analytics.

Click fraud
As well intentioned as the best ad network can be, there is currently no financial incentive for them or other industry players such as publishers to prevent click fraud.  Don’t forget that there are people out there who commit click fraud and make money from it. I’m talking about things like stacking adverts, where six or so ads are layered on top of one another, and when a user clicks once, six clicks are racked up. Money can also be made from running many computer-based simulation programs in parallel, to simulate hundreds of phones. Some networks don’t even check for clicks from simulations.

Another scenario is where adverts are placed or hidden in a game so that the user unintentionally clicks on them. Pyramid Run, an app game, is a classic example of this. To play the game, the user taps the screen to make the man jump and run. Ads pop up spontaneously in the middle of the screen. Aside from causing user frustration, it has also succeeded in clocking up a huge number of false clicks that the advertiser would have had to pay for. Pyramid Run was pulled from the Apple App Store due to copyright issues with the popular Temple Run game.

But metrics go beyond clicks and impressions. As ad analytics in the online world have matured, the in-app advertising industry still has a long way to travel and many technical hurdles, but there are solutions out there.

As I see it, standards KPIs need to be set for in-app advertising. Impressions and click metrics should take into account time to click and unique clicks to make sure that what we are measuring is valid and useful. Additional basic KPIs should include downloads, conversions, one-time users, retention percentage, average number of sessions, and average revenue per user. The technology is already available to deliver these metrics.

With these data points, marketers can move into user loyalty analysis and cost-per-conversion calculations, making informed decisions easier. Data on country, language and device will also give you a degree of confidence that your mobile ads are heading towards the right targets. Just over the horizon in 2013, brand safety data points will also come into play.

Mobile marketing no longer needs to be an unknown black hole sucking in big budgets. Mobile web guidelines are being finalised by the IAB, and I am confident that early in 2013, in-app advertising standards will be at least be drafted by the IAB for discussion. It is a simple task of implementing smart business models, and expecting valid, standardised and verified metrics to justify mobile ad spend.

Giuseppe Bellanca is managing director, EMEA, at adeven