Seems like people are finally getting it: big data is going to change everything.
From a business context, big data has one goal: to serve as a base for analytics that drive actionable insights to help you make informed decisions for your business and organization. In a general sense, the world of data was simple before mobile devices and the cloud emerged. Previously, most companies’ brand presence consisted of a website and a physical location.
Things have become slightly more complicated. We are operating in a world of multiple mobile devices – tablets, smartphones, and wearable devices. It’s not unusual for consumers to own, even use, more than one of these devices simultaneously.
As a result, this increase in devices generates larger and larger datasets with increasing variables and complexity. How exactly do you deal with that? How do you drive actionable insights from this sea of data? How do you make it easier to access these insights?
As Xtreme Labs’ Data Scientist, here are the strategies I recommend for clients facing these issues.
Integrating Analytics with Big Data Analysis
A perfect example of this is Google’s announcement at its recent I/O conference that it is integrating Google Analytics Premium and Google BigQuery. For the uninitiated, BigQuery is a service that allows users to analyze massive amounts of datasets.
The difference between Analytics and BigQuery is that Analytics aggregates all the data and displays it; however, it doesn’t let you customize the types of data or access deeper levels of analysis. BigQuery lets you go where Analytics doesn’t: you have the ability to access data on a much deeper, granular, level.
This is somewhat similar to Flurry Analytics, which does the aggregation and permits users to download the datasets as well. However, in the case you have high volumes of traffic, there are going to be obscene amounts of data to analyze and trawl through. Even Excel has a maximum worksheet size of slightly over 16 million entries. BigQuery can handle much larger amounts of data, and makes sure that this data is not locked into analytics tools.
Google also announced at I/O that their new sign-on experience – User ID for Google Analytics – would provide a much more accurate count of unique visitors by enabling cross-device reporting. Ordinarily, if a user visited a website from a laptop and then from their mobile phone, she would be displayed as two unique visitors. However, if the user is signed into Google, they will know that it’s the same user and register her two pageviews as one unique visitor, instead of two pageviews and two unique visitors. This type of identification also opens up the ability to perform device path analysis, which looks at the types of devices your visitors are using to navigate your website and allows you to track behavior further.
Convergence analytics brings together disparate sources of data into one dashboard. This dashboard can display a variety of metrics or variables, such as web usage, mobile usage, campaign data, or various other user activities – keyword entries, click-through rates, mobile, and other customer touchpoints. Currently, a lot of this data is in separate silos (e.g., website and mobile metrics are separated).
By freeing the data from its silos, and having a dashboard that enables viewers to see all these metrics at once, analysts can make connections and analyses faster and in a more relevant way.
LinkedIn makes great use of predictive analytics; after gathering a user’s data and applying it to a statistical model, the network is able to make suggestions on who the user may want to connect with based on current information.
Various other examples of predictive analytics exist already: credit scores, Amazon’s recommendations, and even Google Now. This is the next frontier of analytics: tailoring services to consumer’s personal interests and preferences. The large amounts of data that the cloud and mobile provide makes predictive analytics much more precise.
Harnessing the power of digital analytics means you have to look at all the data to make these informed decisions. If you have users coming to your website and mobile app via different devices such as smartphones and tablets, you will want to look at all of that to put each element into context rather than look at each separately.
Nonetheless, we are still in the early stages of this. As wearable devices enter common use, new categories of data and variables such as locations and gestures will become available. It’s time to get familiar with all of these areas, which doesn’t necessarily mean you need to become a specialist in big data, but it does mean understanding what big data can do for you.
If this all seems a bit daunting, first figure out your business goals and what analytics or metrics you’re going to need, what insights you’re going to need, and build a plan out from there. Familiarize yourself with all these different areas. And, in case you need help, you know who to contact.
Connect with Jack on LinkedIn.
About the Author