The challenges of mobile measurement are numerous and familiar: a lack of tracking cookies, fragmentation among devices and platforms, the split between apps and the mobile Web, and a daunting array of analytics tools among them.
In the absence of standardized metrics, a new Forrester report outlines strategies for piecing together different data sources to
effectively track mobile campaigns.
Companies should start by categorizing their mobile “subchannels” including search, display, apps, Facebook and ad networks as a foundation for benchmarking ad performance. Each channel yields its own data set, and first-party data from owned sites and apps are the most reliable.
Layered on that should be contextual data according to factors such as the operating system, handset, location, and time of day, which provide ways of targeting users at the right time. Also, companies should collect interaction data including impressions, clicks and conversions.
The data should be tagged to specific product “life cycle” stages.
“For example, display
impressions connect to ‘discover,’ while Facebook time-spent type data on your page should be tagged to ‘engage,’” states the report authored by Forrester analyst Anthony
Mullen. This approach, he noted, helps marketers see what levels of reach and engagement they are achieving across different mobile ad formats like display, social and branded apps.
Beyond taking these initial steps, Forrester also recommends more specific tactics -- such as turning to agencies with mobile expertise, including SapientNitro, Tribal DDB, and AKQA that have created “custom integrations to draw the disparate pieces of mobile measurement and campaign planning puzzle together,” according to the report. (Forrester ranked top mobile agencies in a previous study.)
The research firm also suggests working with agencies and mobile analytics providers, such as AdMob, Rosetta and Bango, that have relationships with wireless carriers to help identify unique users more accurately. They can also report on which devices and operating systems people are using.
Because of the limitations of Web cookies for mobile tracking, the report advises turning to proxies. That approach itself has become more challenging, given Apple’s restrictions on use of its unique device identifier (UDID) by third parties for tracking app usage. Apple is expected to include a more privacy-friendly UDID alternative in the latest update to its iOS platform.
By building “hybrid” apps that incorporate HTML5 technology, brands can also more effectively collect data from cookies. Forrester notes that Facebook, British Airways and many other companies are starting to take this approach with new apps.
The study stresses the importance of ongoing A/B and multivariate testing around measurement data to ensure that campaigns are optimized for the right channels, device types and time of day. As an example, it pointed to an SMS campaign that MTV ran with Starbucks that tailored messages by handset model and stored that information to retain the right preferences.
With continued testing and refinement, it found the optimum message length to be 120 characters -- even less than the 140-character maximum for a Twitter post.