“How did traffic from search change year-on-year in June?”
In the coming weeks, Google Analytics users will be able to ask questions like this and have them answered in seconds thanks to natural language processing (NLP) and machine learning (ML) technology Google is integrating into its popular analytics platform.
Google says the technology, some of which comes from Android and Search, will help Google Analytics users “better understand and act on [their] analytics data.”
According to Google, analysts it spoke to indicated that they spend “half their time answering basic analytics questions for other people in their organization.” By making it possible for analysts and business users to use plain English to obtain analytics data, they will in theory be able to get the information they need more quickly, giving both groups the ability to focus their energies on the higher-value aspects of their roles.
Ultimately, Google’s goal is to make “getting answers about your key business metrics…as easy as asking a question in plain English.”
Today, the new functionality, which is dubbed Analytics Intelligence, can answer a number of types of questions, such as those related to basic metrics like traffic and referrals, performance, and trends. Users can add qualifiers like date ranges and ask answers to include percentages.
In addition, Analytics Intelligence is capable of answering questions about user groups. For instance, a retailer could ask questions about conversion rates in specific countries, and an advertiser could ask questions about which paid search keywords convert best.
Of course, while impressive, Analytics Intelligence has a huge limitation: it can’t tell users what questions to ask.
Analytics Intelligence will no doubt help many Google Analytics users who don’t have the in-depth knowledge of the service to answer questions on their own. Google Analytics is a powerful service but getting the most out of it often requires a level of knowledge and experience many companies don’t have. Thus, many fail to maximize the value of the free service.
But Analytics Intelligence won’t necessarily fully address that issue because Analytics Intelligence can’t tell users what questions they should be asking.
This is one of the biggest challenges facing companies today. There is more data than ever, and sophisticated tools for analyzing it, but that data and those tools are of minimal utility if the people using them don’t use the right metrics and ask the right questions. For instance, Analytics Intelligence can answer questions like “Which channel had the best conversions in July?” but the answer might or might not be of real value to the business. Analytics Intelligence doesn’t know. And it can’t suggest better, more relevant questions because it doesn’t know what those questions are.
What’s more, because Analytics Intelligence can’t answer strategic questions like “Which campaign should I invest in?”, it’s important for users to make sure that the ease of the new functionality doesn’t encourage over-production of reports and charts that don’t actually help answer key business questions.
While Google is trying to address this with Automated Insights, which was released last year, there’s still no substitute for starting an analytics query with a business question.
To learn more on this topic, check out Econsultancy’s range of analytics training courses.