A new study sponsored by EMC Corporation found that despite the unprecedented expansion of the digital universe due to the massive amounts of data being generated daily by people and machines (an average of 5,247 GB of data for every person on earth by 2020), only about 0.5 percent of the world's data is being analyzed.
This provides both opportunities and challenges as only about 3 percent of potentially useful data is even tagged. (Ironically, only about 20 percent of all data being generated is adequately protected.)
The two-fold challenge for marketers in the next decade.
If marketing firms and IT departments capture, organize, and analyze too little data, they are more likely to capture big shadows with distorted information. If marketing firms and IT departments capture too much data, most of them will succumb to information overload paralysis.
This is doubly true because the amount of information stored in the digital universe about individual users exceeds the amount of data that they themselves create. Every day, algorithms are generating data about everyone who is active on the Internet that is only as good as the proposed algorithm.
The effectiveness of these algorithms can be appreciated simply reviewing the advertising content being served up on various social networks and ad revenue websites. If the advertising doesn't match your needs or interests, it is very likely caused by an ineffective algorithm that is attached to your data (and then attracts more erroneous data). The potential for advertising dollar waste is tremendous.
Author Geoff Livingston scratched the surface in his recent article Does a Social Score Make a CMO? Forbes attempted to determine the top 20 Fortune CMOs baed on social scoring as opposed to outcome measurements tied to their work. Social scoring is largely useless data, casting shadows that may or may not be true especially because people who work for big brands tend to attract more followers regardless of their contributions or real influence (even within the field).
Ergo, in some cases, it is the brand and not the person attracting attention. Likewise, other people work hard to game the system, either trolling for follow-back propositions or purchasing followers to inflate perception. Big data doesn't know the difference, despite claims to the contrary.
These are only small examples of the bigger problem. The point is that as data seemingly becomes increasing accessible that doesn't mean that its value increases with the volume. Even if Western Europe is currently investing $2.49 (USD) per GB, the U.S. $1.77 per GB, China $1.31 per GB, and India $.87 per GB to manage the digital universe, there are no guarantees that big data can capture more than distorted reflections of the people marketing wants to persuade.
Big data can be useful, but only when it is continually vetted by traditional research.
Even as data measurements improve for anyone hoping to track, research, and predict consumer sentiment, marketers ought not become lazy researchers relying on slivers of data or online surveys. If social networks are an opportunity for organizations to be more human, then big data collection ought to work toward understanding behavior as opposed to dehumanizing population groups.
Some marketers today might be surprised to find that their best efforts to automate and insert keywords (as an example) might produce a measured response in the short term but fail in vetting sessions that involve live customer interviews, face-to-face interactions, or even focus groups. While the challenges are not insurmountable and provide additional opportunities beyond the digital universe, it goes a long way in demonstrating the need to think beyond the obvious, especially when what is "obvious" only accounts for a fraction of a percent, less than traditional methods used to capture.
While the study raised many of the questions above, there is additional information that can be gleaned from it for marketers and IT professionals. For more highlights from the study, visit the EMC multimedia overview page. It presents eye-opening information about how much we don't know vs. what we think we know as well as an even analysis of challenges and opportunities for big data.
This provides both opportunities and challenges as only about 3 percent of potentially useful data is even tagged. (Ironically, only about 20 percent of all data being generated is adequately protected.)
The two-fold challenge for marketers in the next decade.
If marketing firms and IT departments capture, organize, and analyze too little data, they are more likely to capture big shadows with distorted information. If marketing firms and IT departments capture too much data, most of them will succumb to information overload paralysis.
This is doubly true because the amount of information stored in the digital universe about individual users exceeds the amount of data that they themselves create. Every day, algorithms are generating data about everyone who is active on the Internet that is only as good as the proposed algorithm.
The effectiveness of these algorithms can be appreciated simply reviewing the advertising content being served up on various social networks and ad revenue websites. If the advertising doesn't match your needs or interests, it is very likely caused by an ineffective algorithm that is attached to your data (and then attracts more erroneous data). The potential for advertising dollar waste is tremendous.
Author Geoff Livingston scratched the surface in his recent article Does a Social Score Make a CMO? Forbes attempted to determine the top 20 Fortune CMOs baed on social scoring as opposed to outcome measurements tied to their work. Social scoring is largely useless data, casting shadows that may or may not be true especially because people who work for big brands tend to attract more followers regardless of their contributions or real influence (even within the field).
Ergo, in some cases, it is the brand and not the person attracting attention. Likewise, other people work hard to game the system, either trolling for follow-back propositions or purchasing followers to inflate perception. Big data doesn't know the difference, despite claims to the contrary.
These are only small examples of the bigger problem. The point is that as data seemingly becomes increasing accessible that doesn't mean that its value increases with the volume. Even if Western Europe is currently investing $2.49 (USD) per GB, the U.S. $1.77 per GB, China $1.31 per GB, and India $.87 per GB to manage the digital universe, there are no guarantees that big data can capture more than distorted reflections of the people marketing wants to persuade.
Big data can be useful, but only when it is continually vetted by traditional research.
Even as data measurements improve for anyone hoping to track, research, and predict consumer sentiment, marketers ought not become lazy researchers relying on slivers of data or online surveys. If social networks are an opportunity for organizations to be more human, then big data collection ought to work toward understanding behavior as opposed to dehumanizing population groups.
Some marketers today might be surprised to find that their best efforts to automate and insert keywords (as an example) might produce a measured response in the short term but fail in vetting sessions that involve live customer interviews, face-to-face interactions, or even focus groups. While the challenges are not insurmountable and provide additional opportunities beyond the digital universe, it goes a long way in demonstrating the need to think beyond the obvious, especially when what is "obvious" only accounts for a fraction of a percent, less than traditional methods used to capture.
While the study raised many of the questions above, there is additional information that can be gleaned from it for marketers and IT professionals. For more highlights from the study, visit the EMC multimedia overview page. It presents eye-opening information about how much we don't know vs. what we think we know as well as an even analysis of challenges and opportunities for big data.