Ordinary Web analytics have been with us for years. The science and/or art of measuring on-site engagement has become an article of faith for marketing entities worldwide. Arguably, no other medium has more potential for direct feedback on a marketing campaign’s effectiveness and return on investment.

For magazine publishers, the value proposition of analytics has been less clear. In the brave new world of free Web content, knowing the number of page views, unique visitors or time spent on a particular page has not always meant increased ad revenue. Banner advertising has been particularly hard hit, as advertisers discount CPM rates and become increasingly skeptical about certain data as a real-world measurement of engagement.

Part of the problem is the sheer volume of the data. Even when savvy marketing resources are deployed, however, publishers and advertisers worry about the practical (i.e., profitable) benefits of analytics, given revenue declines on the print side.

Upwardly Mobile?

More recently mobile analytics—particularly for tablet devices—have altered the landscape for publishers. Unlike the Web, mobile/tablet publishing represents a more controlled environment, emulating the print experience but adding interactivity and potential reader engagement in a self-contained vehicle: the mobile app. In the tablet world, a reader’s attention can, in theory, be more focused on content of the publisher’s or advertiser’s choosing. Measuring behavior in a tablet app—or even an app-like HTML5 equivalent—is far more valuable, potentially, than its Web antecedent.

Tablet analytics do share many of the features of mainstream Web analytics. Data about devices, operating systems and other technology used to access content include the device brand name (e.g. Apple iPad), OS version, screen size, carrier information and the presence of support technologies like Flash—which, in the Apple example, would be “no.” Google derives device information from its own database, while others, including Omniture, rely on services such as Device Atlas. More esoteric—but still essential—data would include bandwidth detection (e.g., 3G vs. WiFi).

Technical data is only the beginning, of course. Tablet analytics include such behavioral metrics as page views (arguably more significant for a print-like tablet page than an unconstrained HTML “page”), time spent, bounce rates, user actions (e.g., clicks), specific article metadata, ad responses, churn rates and the like.

What distinguishes tablet from Web analytics centers on the nature of a mobile device itself. Location in particular is a metric of particular value to publishers and advertisers. Standard Web analytics can muddle along with geolocation, using IP addresses, MAC addresses and RFID, but a mobile device inherently “knows” where it is—providing data that can benefit both the consumer and advertiser. If a user grants permission to use a device’s location, a tablet app can provide extremely useful data, even to the point of customizing ads on an individual basis. Used wisely, geolocation data can give proactive publishers and advertisers the potential to actually serve their readers’ interests and build formidable brand loyalty. (Used unwisely, however, location data will generate a whole new class of digital annoyances.)

Other tablet- or mobile-specific data include the specific capabilities of tablets and their unique properties apart from geolocation. Multi-touch navigation, as well as compass and gyroscopic functionality, will undoubtedly create a whole new class of analytics. In the long term, this will not be limited to the gaming industry, since game playing is already a typical attraction for Web and tablet advertising.

Where tablet analytics falls short of its Web counterpart is related to search and traffic source data. When the native app model is used (as opposed to HTML or HTML5 in a mobile browser), classic Web analytics regarding search engines and external site links are problematic. However, as HTML5 and CSS3 enable developers to replicate the app experience in a browser, this limitation may become moot.

Java on Board

Mobile analytics emerged as handset manufacturers and carriers began adding Internet connectivity to their offerings in the early 2000’s. This was not an easy task. Traditional on-site Web analytics often relied on HTTP cookies or JavaScript on the client side. However, early browsers in not-so-smartphones were extremely limited affairs.

More recently, smartphone and tablet browsers are increasing in overall capability, including better implementations of JavaScript. As a result, mainstream analytics companies have begun to directly support the mobile world. This includes more traditional device and usage data, as well as the highly desirable geolocation data coveted by advertisers. However, skeptics believe that JavaScript-based analytics are still problematic, performance-wise. Until the tipping point arrives, the client-side approach will be problematic for straight HTML5 or XML mobile analytics.

For publishers, however, mobile browsers are not the only game in town. Native apps now dominate the publishing landscape for smartphone and tablet users. Apps provide not only a controlled vehicle for content delivery, but also a built-in infrastructure for generating useful behavioral data.

Tag, You’re It

On-site analytics often relies on page or element tagging—the insertion of code that, when executed, indicates a specific user action, such as opening a page, clicking on an object, entering information, leaving a page, starting or stopping a video playback, and so on. This user-generated data, combined with automatic data about environment, location, etc., must be interpreted on several levels, in order to provide meaningful information to marketing analysts. Therefore, the design and implementation of these tags is a critical first step.

For HTML- or HTML5-based content, tags expressed as JavaScript or a comparable framework is a well established practice. Where content has not been expressly tagged, many vendors parse the HTML file itself, passing the content of <h> tags, for example, into the stream of analytics metadata. More advanced systems utilize already-tagged content from a CMS, giving publishers and advertisers a more detailed look at what is being viewed. No matter how sophisticated the environment, however, the axiom “garbage in, garbage out” applies with a vengeance. If the HTML structure or the CMS approach is poorly planned, the quality of the analytics will suffer.

On the app side, tagging follows the same methodology, but with more environmental variations than straight HTML-based tagging. Each mobile operating system has its own SDK, dictated by the characteristics of supported devices. Content tagging within apps is comparable to its browser counterpart—mainly because it is expressed as HTML5—tablet-specific environment, action and location data is not. This means that analytics developers today must create a tagging infrastructure for iOS and Android, at a minimum, plus Tablet OS (RIM), WebOS (HP) and possibly others, as well as conventional JavaScript. The need for standards would seem to be obvious, but even conventional Web analytics has been resistant to unified standards efforts to date.

Despite these difficulties, most of the scores of traditional Web analytics developers are eager to expand into the mobile arena, according to Phil Kemelor, VP of the analytics consultancy Semphonic. As the use of smartphones and tablets increase, vendors are discovering how to capture meaningful device and usage data from native apps, and help publishers interpret that data into meaningful, actionable information.

Paradoxically, Kemelor believes that publishers are currently less interested in mobile analytics than they are in simply getting their content onto mobile devices. For now, this means that vendors are the primary drivers of what metrics should be tracked. That will soon change, however. Kemelor believes that the adoption cycle for mobile/tablet analytics will be much faster than it was for conventional Web analytics.

Meet the Vendors

Of the many vendors in the general Web analytics field, several stand out in the area of tablet analytics for publishers. (The field itself includes a number of offshore companies, as well as a considerable amount of consolidation and acquisition.) A few publishers are engaged in self-development, but most rely on analytics companies identified by developers of tablet publishing environments, such as Adobe (DPS), Mag+ and WoodWing.

Recently acquired by Adobe, Omniture offers a broad range of Web and multi-channel analytics tools including, not surprisingly, support for tablet apps created in Adobe Digital Publishing Suite. (Other tablet app developers, notably WoodWing, offer Omniture integration as well.) For publishers, options range from a “light” version of basic environmental and usage data to more advanced fare, tracking detailed behavior and integrating it into a marketing intelligence environment.

Ray Pun, senior product marketing manager for mobile at Omniture, commented that the company’s multi-channel approach has potential value outside the tablet sphere. User authentication and a consistent user ID, for example, can allow users to track behavior on other media—albeit indirectly, via surveys, with non-digital media.

Another software giant, Google, has long offered its Google Analytics service for general Web developers, as well as mobile tracking SDKs for Android & iOS applications. GA also has third party tracking libraries for Flash / ActionScript, Adobe AIR, and Microsoft Silverlight. Last month, the company announced enhanced mobile support with reports like detailed device information, including images of each device, as well as an array of use and geolocation data.

Phil Mui, group product manager for GA, described the company’s success with several major publishing projects, including the New York Times’ AIR-based Reader. For mobile web sites tagged with GA, he said, the publisher can rely on GA to track content consumption. User actions in a mobile app can trigger measurable data, which include metadata derived from a CMS or parsed from HTML. As could be expected, GA is integrated with other Google products, including AdSense, which can deliver ads customized by location and other contextual data.

Smaller players have also made significant progress. Boston area-based Localytics has an impressive list of publishing clients, including The New York Times, Dow Jones, Meredith and National Geographic. Software partners include Mag+ and (more recently) WoodWing. Although the company does provide JavaScript tagging for HTML5, its main strengths include support for application tagging in iOS, Android, Blackberry and Windows Phone 7. In addition to a full set of device and use data, Localytics’ SaaS environment provides advanced e-commerce funnel metrics and critical data about the preferences of those who follow the publisher’s brand.

Localytics’ Market Development Brian Suthoff believes that publishers are embracing analytics much faster than was the case with standard Web analytics. Since mid-2010, he maintains that publishers have switched from simply rushing content to mobile platforms to viewing analytics as part of their sustainable business practices.

Another developer, U.K.-based Bango, is focused exclusively on mobile/tablet analytics. According to senior marketing vice president Anil Malhotra, the company’s offering is entirely server-side (apart from native mobile browser or app functionality), providing a long list of device, network type and user behavior data geared to the needs of mobile-savvy advertisers. Measuring known user characteristics and behavior has enabled Bango’s publishing and advertising clients with the ability to optimize their campaigns over time.

Planning for Success

Every vendor we interviewed was clear on the basic strategy and approach publishers need to follow. Advance planning and careful implementation of a tagging infrastructure are high priorities, with the goal being that analytics are built into the workflow. Editors and production staff should seldom if ever be concerned with tagging content. While publishers will benefit from top level data, the consensus is that analytics data, by itself, requires marketing expertise and interpretation. As the tablet channel grows, publishers will have the data to create a more sustainable business, provided they know how to use it.