Google Analytics: May I Suggest This?
- by Laurie Sullivan @lauriesullivan, September 25, 2009
The tool may highlight the information needed to improve conversions, rather than serve up all the raw data that cloud decisions. Today, the Google Analytics dashboard presents all the data in columns. In the future, the tools could suggest the most important data that requires attention. "It might alert the user when conversion rates in the United Kingdom decline," says Avinash Kaushik, Google analytics evangelist.
Confident that Google engineers have the knowledge and the mathematical skills to build the algorithms, the tool will provide more intelligence and less data, shortening the gap between the access to information and the ability to take action. "It's the ultimate quest for every analytics provider," Kaushik says. "We're very passionate about not doing data puking, which is essentially what many tools do today."
When the conversation focused on data from real-time search, Kaushik said Google is getting closer to providing data from near-instantaneous searches. "This week I ran queries on Google News," he said. "They were two hours old or less, and that's pretty darn close to real-time."
Kaushik says he picked up a URL on Twitter that allowed him to access data in a search across some Google properties. The data had run across the network within the past hour, giving him near real-time access to search data that could tie to paid search ads. Marketers could also theoretically gain access to more real-time data sets.
Marketers have access to a ton of data, but not many know what to do with some of that data. Knowing something happened at the moment it happens provides benefits, but not being able to do something with the data presents a problem.
For example, Insights for Search lists the most rapidly rising terms in the report. The words start at the bottom of the list and rise to the top. That makes data more actionable, Kaushik says. The person using the tool doesn't have to mine through tons of data to get to the important information -- it's done through algorithms. There's no limit to the information people can mine on the Internet, but tons of limits as to what they can do with it, which makes the statistical significance more important, Kaushik says.
But that gap is shrinking. Google engineers across the globe, from the United States to Israel, work on approximately 300 experiments daily related to data information, making marketing more efficient and producing algorithms that can react to real-time information. The experiments range from improving search results and maximizing the return on investment between search and display ads to advanced bidding algorithms.
"The greatest gift you can get from the Internet is the ability to fail faster," Kaushik says. "Today, if I create a failed magazine cover or offer on TV, it would take me weeks or months to figure out I was wrong. The price of failure is high."
Kaushik can figure out within 24 hours if he has made a mistake on something that appears on the Web. He calls it "liberating" -- the ability to take risks, quickly discover failures and correct them. The company that understands the Web allows projects to fail faster and gains the knowledge to succeed in finding new business models, he says.
