Can Facebook replace daily horoscopes? Given the hoards of social graph data in its warehouses, a social media network like Facebook could very well have enough information on users to provide useful recommendations on a smattering of subjects. From romantic pursuits to career endeavors, Facebook alone has been documenting our lives for over a decade.

As part of my ongoing research regarding social media’s potential as a data-driven horoscope replacement, I spoke with TIBCO Software Inc.’s Brian Gentile. As the senior vice president and general manager of TIBCO Analytics, Gentile was especially insightful on the world of data visualization. Gentile recently shared his thoughts on social media’s business opportunities in delivering useful recommendations, what that would look like and if users would even respond to such a service.

Data-driven horoscopes a real possibility

Q. Do you think individualized data reports can replace human trust in less scientific predictive methods such as horoscopes?

Gentile: Creating a summarized “personalized newsletter” or daily briefing should become a feature of most social networks. Because it would be based on authentic data (or at least data built on status entries of all of those in one’s network), it would have the ability to draw correlations, synopsize many items into useful social information, and even predict the likely events in the coming day (or week). In this regard, we may well have enough information for a data-driven horoscope.

Q. Are there any correlations you’ve encountered between astrology and advanced analytics?

Gentile: No specific correlations yet, but I keep a very open mind. I believe that our ability to predict the likelihood of future events will greatly increase during the next few years. The impact these predictions might have on our personal lives is only beginning to be considered.

Q. Given the amount of personal data social sites like Facebook collect, do you think these social platforms could one day offer a recommendations service akin to daily horoscopes?

Gentile: I think that a combination of Big Data and advanced analytics will provide something even more personal, accurate and useful than traditional horoscopes. Based on readily available data from one’s social networks and geo-position, combined with consumption patterns, commute routines, travel plans and so on, an accurate, personal “serendipity index” could already be formulated. For example, considering all that is known about an individual and his or her preferences and plans combined with the social graph, a prediction could be easily offered: “Dress sharply: Today you will meet someone important you’ve not seen in a long time.”

TIBCO Recommendations

Data-driven recommendations … don’t annoy the user

Q. What real-world use cases have you encountered of data-driven recommendations designed for average consumers?

Gentile: The entire concept of recommendations is based on data-driven engines that correlate many useful bits of data (gathered mostly through the clickstream, personal profiles, and so on) and then use those bits to suggest reasonable, relevant alternatives for one to consider. Nearly all advanced, online experiences today incorporate data-driven recommendations. Examples include dating sites, commerce sites (Amazon invented the commerce-oriented recommendation, and it has been hugely successful), and sites that help anyone better manage their money.

In each case, an “average consumer” is made much more capable as the metadata they are providing during their search and exploration is put to use uncovering and suggesting other relevant possibilities in nearly any topic of interest. My personal favorite is the online investment recommendations suggested by my financial services firm, which have helped me to be a much better-informed investor. This is a growing trend, where marketers, retailers and insurance companies, in particular, are already using data to improve customer interaction and end-user experiences.

Q. Do you think this will be the primary way in which data is applied to consumers, or do you think an actual consumer-centric product could survive?

Gentile: What’s usually lacking in most of these applications consumers engage with, whether banking or retail, is context. So presenting data to a consumer in one of those environments, if done in the right context, could be helpful. There’s an opportunity here. An enhanced recommendation engine from Walmart could be helpful — for instance, if they say 74 percent of people in my age category that bought product A also bought product B & C.

The less relevant it is, the more negative the experience. Consumptional context is everything. And you can apply that to every industry. Education will be different than a dating site versus a banking service. It takes clever designers of these tools and services to make it usable and relevant. We have the potential to use data in a way that’s helpful, but there is also the potential to annoy the user.

The art of visualizing data

TIBCO Software Inc. recently released its Recommendations tool for its Spotfire Cloud platform. The tool allows everyday business users to visualize analytics insights using charts, graphs and widgets, making it faster and smarter to make critical business decisions. Recommendations uses a built-in analytics intelligence wizard to create best-practice visualizations in just a few clicks.

“Recommendations provides a fast and intelligent jumpstart to the analysis of data by business users, helping anyone to select the perfect visualization to discover insights and tell a story with their data, enabling them to focus more on guiding business decisions and less on selecting the ideal chart, filter and perspective.”

Q. How do TIBCO products incorporate recommendations?

Gentile: TIBCO’s tools offer a wide set of choices for charts, widgets and graphs to visualize data, but the user has to select it. That’s why we made the change to offer it through a recommendation because it’s so powerful, and it’s the first tool in the world able to do this. When you shop at Amazon, for example, it’s well known for making recommendations based on browsing history. Its revenue relies on this. Its shopping service gets to know you well enough to recommend the products you want or may have had trouble finding otherwise. We give users a starting point to look at data in ways they may not have thought of themselves. With Spotfire, we’ve built a set of recommendations. As soon as it ingests the data, it can immediately offer a visualization to begin seeing the pattern in the data.

Q. How does TIBCO help contextualize collected data?

Gentile: The primary question is how do we help more people who are less skilled than a data scientist to act on data they would’ve otherwise missed? We have a real responsibility today to put a lot of data to work because it’s easy to do. The concept of data being presented at the right time is as part of the context. Fast data makes big data small. It lets you get to the right data at the right time.

Feature image by manoftaste.de via photopin cc