From KGB to IOT, with love

There was a time during the Cold War when the KGB was so good at detecting CIA agents infiltrating the East as mere State Department employees that the Americans thought they had a mole in their midst. According to a article written by Jonathan Haslam, a professor at Princeton University, the Russians' source was somewhere else entirely.
It seems that the Russian service used early data mining techniques. The KGB did nothing more than collect public information about US Foreign Service employees and compare it with the information it already had about exposed agents. Putting the data into context highlighted some differences in the lifestyles of various State Department „employees” in the East. Thus, the Russian service discovered no less than 26 independent indicators that differentiated CIA agents from ordinary American employees.
For example:

• Covered agents were significantly better paid.
• Regular employees usually returned home after 3-4 years. Agents did not.
• When agents returned home, they did not appear on the State Department's public lists.
• Regular employees were recruited before the age of 31. Agents could be older.

So, no high-level mole, just outlining patterns in data obtained from multiple sources. Read-Write website compare the example of data analysis done by the KGB with the information that can be obtained in business by analyzing a "pile" of data enterprise. The publication applied the KGB model to marketing, to show that the equivalent of indicators used in Cold War analyses can be discovered in consumer data:

• Which consumers leave after the product launch period and which renew their services?
• Which customers upgrade their products, and which switch to another brand?
• Which consumers buy directly online and which buy in-store after researching online?
• Which consumers are motivated by a discount and which are only interested in certain features?

Just as KGB analysts used data on known CIA agents to uncover others, companies can extrapolate certain patterns to apply to the marketplace. A particular age group, gender, geography, purchasing and payment method, etc., can signal customers on the verge of leaving or upselling opportunities.
Read-Write offers some examples:

• An online retailer can analyze online traffic to differentiate between uninterested visitors and those whose behavior suggests they could become buyers, if stimulated a little.
• the same retailer can segment its marketing strategies on seasonal or permanent customers, to address a Christmas customer at exactly the right time.
  

There is a famous case of the American teenager who received promotions on newborn products from Target, until the girl's father, indignant, complained to the store. Target apologized, but what the father did not know is that the young woman was indeed pregnant. The data-marketing department of the American retailer had intuited this by analyzing the purchases she had made in recent months.

Analyze this.

It is clear that extracting meaningful insights from disparate data was not invented with the internet, it was just automated and moved into the business realm. Things will get even more exciting with the rise of the next wave of personal data, driven by the Internet of Things. IDC shows that investments and spending in IOT will grow from $699 billion in 2015 to almost $1.3 trillion USD in 2019. If in 2015 the impact of IOT was felt most in manufacturing and transportation, in the next five years we will see the greatest growth in insurance, healthcare and the consumer market.
Everywhere, sensors connected to the internet.