Category Archives: Prediction

Link

The real trend this year is not technology. It’s about helping business people to make better decisions, and actually change the way companies are doing business. How can big data be used? The curtain raiser 2012: The Year Analytics Means Business lists that:

1) Red Bull has been able to speed up and simplify their data warehousing environments.

2) Colgate-Palmolive, Provimi, and Danone use vast data amounts to optimize their production and sales.

3) Medtronic uses data to access and analyze the large amount of complaints and feedback data they receive about their products, combine it with other data sources, and provide it to business users through dynamic interfaces.

4) Houghton Mifflin Harcourt uses sophisticated, risk-based forecasting. The company prints books in January or February, when printing is much cheaper. In order to minimize excess inventory, it has carefully analyzed all the causes of previous forecasts, and now takes account of all the different facts that influence book obsolescence.

5) International grocery chain Casino is rolling out a new mobile shopping application for its customers. It provides data from its enterprise systems directly to its customers, resulting in increased shopping convenience and increased customer loyalty.

It is a good read by Timo Elliot (@timoelliot). Note: it is a very SAP-centric article.

Link

Social and mobile measurement and insights are about to bring marketing measurements to a new high. The following examples are given by Alex Cocotas and Henry Blodget (@hblodget) in a Business Insider presentation:

  • Rent the Runway has 200% higher conversions from referrals from fashion magazines then from paid searches.
  • ShoeDazzle says Facebook connected users that are 50% more likely to make monthly repeat purchases
  • Friends referred by friends at One Kings Lane have 2x lifetime value of customers from other channels
  • When a Facebook user clicks on a Ticketmaster purchase shared by a friend, it is worth between 6 and 8 $ in new ticket sales

See the full presentation: rise of social commerce

Police are Using Big Data To Prevent Crime

[vc_row][vc_column width=”1/1″][vc_column_text]

image

Did you enjoy the minority report?

Whoever enjoyed Steven Spielberg‘s vision of crime protection in “Minority Report” can look forward to the real thing now. UCLA mathematician George Mohler uses seismologists models for crime predictions. So far, the US police have mainly identified hotspots where crimes happen often. Inspired by recent innovations in big data and by businesses like Walmart and Amazon, police are trying to be predictive.

Past crimes can be predictive of future crimes.

The numbers speak for themselves: officers have reversed the crime wave and lowered crime by 11% from the first half of the year to 4% below historical averages for those months. The full article is here: Predictive Policing with Big Data. Those numbers are impressive, but as usual, a word of caution should be applied. It is often not easy to see what is causality and what is not. As Freakonomics showed with their argument that abortion reduced crime rates, it can be difficult to differentiate between different factors. Only once this model is tested more going forward can one attribute crime reduction to Mohler’s algorithm. Despite this, these developments show how fast we are moving towards a more predictive society.[/vc_column_text][/vc_column][/vc_row]