When we saw that Google TV was released to the public 6 days ago, we thought: what kind of impact might this have on web analytics?
First of all, what is Google TV? Basically, a partnership between Google and Intel, Sony and Logitech to essentially turn any HDTV into a media PC. One of the goals with Google TV is to enable new innovation from content creators, programmers, developers and advertisers. By bringing access to the entire Internet, the idea is that you can now easily navigate to thousands of websites to watch web videos, play Flash games, view photos, read movie reviews or chat with friends—all on the big screen.
1. Difficult to Differentiate TV Traffic from other sources
First potential problem: can we tell which visits are from TVs? They might be identified differently in the operating system description, but they may appear identical to any other device on other critical aspects. For example - although TVs are larger, they have pretty much the same resolution as most computer monitors (and even some mobile devices). They use full Chrome browsers, capable of running Flash or anything else. They run Android. The connection speed is identical to a regular computer.
2. Warping of visitor metrics
When visits to your site are from TVs, measuring any kind of "visitor" metric gets more complicated.
First, TVs are only one of the many ways that people will use to access the Internet. And they will probably not be the primary means of access. It is safe to assume that a visit from a TV will probably be accompanied by another visit by the same person using some other device. It's near enough impossible to measure things like "unique visitors", "new visitors", "time since last visit", etc., if the average visitor uses several devices to access the Internet on a regular basis. So all of our measurements will need to focus more and more on visits, as opposed to visitors.
Secondly, TVs tend to be used by groups rather than a single person, unlike computers and phones. That means one visit from a TV may in fact represent several people, so the definition of "visitor" becomes more and more uncertain.
3. Conversion Attribution
Knowing which traffic source converted will become much more difficult to analyze. Imagine: a person is introduced to your site while using Google TV with other people but decides to go back later on their own PC. How do you attribute the conversion? It is impossible to connect the conversion to the initial visit from the TV. And therefore it's also not possible to compare number of conversions to the number of people who viewed the ad on TV.
What's more, how is conversion measured when a visit is triggered by one member of the audience who has visited your site before? Is it a new visit? Is it direct? A referral? If it was found using a search engine, but after first having being recommended by an audience member, how is this going to be recorded? This isn't really much of a new issue in web analytics, but is likely to become more of an issue with the use of Google TV.
This could open up new types of debate - not just where and which devices do visitors come from, but what order these devices are used and how many are used before a visitor finally converts. This will enable us to draw better conclusions about how and why particular devices are used.
4. Group Dynamics
With TV audiences, groups have very different navigation patterns than individuals when browsing.
The group will always exert some influence on how you interact with the site. What you do online with a TV among a group of people will reveal more about group dynamics than individual tendencies. Decisions will probably be less whimsical and more critical. They will also be more likely to favor cultural or moral norms.
The goals of a group will be different to that of an individual. They will be less likely to make a purchase, for example, unless it involves the whole group. Pages will have to be optimized for groups instead of individuals.
5. Separate Apps and Sites for TV
Google TV is unique in it's focus on user-generated applications. Starting in 2011, the Android Market will carry apps designed specifically for TVs. This will shape the way TVs are used and how we think about them. Using the web on a TV will no doubt produce a few limitations and opportunities for websites, so it's a given that websites will begin to take into consideration building versions for the TV, just as they build mobile sites for browsing on a mobile phone.
What Does Google TV Mean for Web Analytics?
This list might be disturbing for web analysts. Google TV and other emerging technologies will change the way we perform web analysis. Just when we are beginning to become familiar with what different numbers represent, those numbers either become irrelevant or change meaning altogether. But this not mean that web analytics is void, nor does it make it less reliable. It just means that web analysts and companies who are online in any way will need to be smarter about how they interpret reports. One way to do this is by using a tool on top of their GA data, which allows them to segment and dig deeper into the numbers.
The cool thing about this industry is that nobody has determined "the best way" to measure activity online. The debate over what to measure and how to measure it is ongoing. The emergence of Google TV and competitors will keep good web analysts on their toes and add a new voice to the debate.
(topic inspiration from Analytics Market)