Finding a Workaround for Effective Mobile Measurement

25 February, 2016

Despite common beliefs, the technology required to measure the return on a mobile ad investment (particularly within apps) is fairly robust nowadays. A wide range of suppliers can now measure any action taken by the user on a mobile device after seeing an ad. And yet, mobile measurement remains a challenge for the industry.

Mediasmart tells ExchangeWire that the standardisation of mobile measurement tools is required for advertisers to keep up with the pace of mobile influence.

The ability to source accurate measurements comes with the caveat that various technical challenges exist around measuring the ROI that can be directly attributed to a mobile campaign. As is the case with all media types, these technical challenges present themselves when an advertiser is trying to measure attribution across channels. As an industry, we need to look at performance of campaigns across all channels, whilst also considering the influence of each channel over the others, while at the same time, addressing the “if you didn’t measure it didn’t exist” mentality. But this not a challenge that is specific to mobile measurement.

From basic ad viewability to user engagement that justifies ROI, the main challenges to overcome are actually non-technical ones. Whilst the tools exist, often they are not considered “reliable” by the big players. This is largely because the technology required for proper measurement is not used by the vendors that have established themselves as the standard in the digital advertising world.

The fact that the majority of measurement tools available are desktop-first solutions compounds the challenge in the current justification of mobile ROI. The technical nuances of the mobile ecosystem and, in particular, of the in-app inventory, mean that desktop orientated tools focused on web have a long way to go to be truly fit-for-purpose. Until they do develop sufficiently, or mobile-first tools become trusted by the agencies and advertisers with the big budgets, mobile investment will continue to lag behind its full potential.

So what’s the best workaround for now? The obvious way to justify ROI is by measuring conversions, engagement and viewabilty etc., but here’s the main caveat: To accurately measure ROI on a mobile campaign and avoid wasting advertising money on mobile, the first step is to be aware of (and then compensate and account for) some of the issues inherent in measuring with desktop-first solutions.

The inaccuracies with current mobile measurement solutions often stem from the difference between desktop and mobile devices. Where advertisers and tracking tool providers are measuring website engagement that results from a mobile advertising campaign, for example, there are a couple of differences that could generate huge measurement inaccuracies and discrepancies.

Cookies and Fingerprinting in Mobile

Cookies don’t work well in the mobile ecosystem. In the mobile advertising world, a customer increasingly clicks on an ad within an app, so the destination URL is not necessarily opened by a browser on the mobile device, but inside the application. The destination URL is opened via a custom browser within the app (or web-view). When the user closes this web-view, he will stay within the app as he navigates through the website.

One peculiarity of this web-view is that each application is completely isolated from the rest, and also isolated from the browser in the mobile device (i.e “sandboxing”). This is a fundamental aspect of the OS security for Android and iOS.

The downside is that sites that open within applications this way will not be able to read or write persistent cookies. So desktop-first tools will not be able to use cookies to identify whether a user or a visit is new. In some cases in which persistent cookies cannot be written, some tools are designed to use combinations of IP address and user agents in order to track. However, this “fingerprinting approach has a couple of important caveats as well:

1. The full user agent in mobile devices is not always usable, and often the information taken from the user agent is limited to the model and make. Also, devices such as the Samsung Galaxy and iPhone/iPad present little or no fragmentation when it comes to user agent. Given the success of these brands and models, many clicks may have taken place in devices that have identical user agents.
2. The IP is not often usable either, because on mobile, many devices will share the same IP while they browse on the mobile carrier network, and secondly, most mobile customers use a shared IP model that originated with the old WAP gateways.

In a nutshell, campaigns that have served impressions and generated clicks in web-views within apps, and are tracked with systems that are based on cookies, will not measure users or corresponding visits. Even if those systems have a fallback fingerprinting method in case cookies can’t be used, they will not measure many users or visits that are served in Samsung and iOS devices via 3G connections. The situation would also be similar on iOS for mobile web of course, as third party cookies are prohibited.

Effective mobile ad measurement to justify ROI in a particular campaign is a question of trusting the right technology providers. They do exist, even if they aren’t mobile specialists.

However, there’s a long way to go until the industry will be able to use these tools in a way that allows commitment to ROI guarantees and viewability standards – and to get there the various industry players need to be aligned. This can only come through standardising technical solutions for the in-app environment, and through industry maturity, as conflicts between advertiser and publisher goals for viewability measurement has shown.

Within this chicken-and-egg situation, it’s likely that the demand side is best positioned to take the lead towards a robust standard. After all, it’s the DSPs who are able to measure whether an ad has been viewed or not and to systematically optimise a campaign to place more value on those offered placements more likely to become viewable impressions.

Meanwhile, as an industry we all need to figure out how to approach this alignment so that measurement tools operate in the interests of everyone – and we need to do this soon. With the growing pace of mobile influence, accurate viewability guarantees and ROI measurement will become imperative to the very existence of the mobile advertising industry.

It’s clear that both the required technology and industry standards will come together eventually. What remains to be seen is whether the desktop first players step up to the plate to provide this, or whether existing and new mobile specific offerings rise to gain the trust of the advertisers, in the same way as they did for in-app tracking and attribution – and it’s even bet either way.

Other Recent Work