This conversation is closed.

What would an analytic model for the humanitarian world look like?

Google does it. So does IBM. Target has been able to predict when women are pregnant, sometimes even before they know it, with this tool. How do we harness it for the good of the world?

Analytics is kind of a buzz word in the corporate world today. Many people are talking about it and excited about the possibilities it offers. With this tool you can simplify a vast amount of information into a bite sized pieces that can be tailored to apply to anyone.

How it Works
By no means am I an expert in the realm of analytics but I know a little bit about it. The idea behind business analytics is taking all the information from a particular customer and tracking when they buy certain products, what they buy, where they buy it, and they try to find patterns in these habits. This is the theory behind Google's analytics and even Facebook's admin section if you ever own a page. It is just mass amount of information that is filtered into a form which can almost predict a customer's needs and wants. This is good for the customer and the company because they both get what they want.

Well this theory is all well and good for companies and corporations who can track buying habits, but how do we do this for non-profit organizations who don't have "customers" per say but does have people they serve and have specific wants and needs?

The question is this:
First, what information would humanitarian agencies be looking for? (weather patterns, amount of people in each county, number of jobs open, apartment available in a region and prices for incoming refugees)

Second, depending on the topic, how would we quantify some of the information out there that would be needed?

Or am I asking the wrong questions? Should I be asking more? Basically how can this idea of analytics be transferred to humanitarian needs?

  • thumb
    Aug 30 2012: Is this the sort of thing you are talking about?
    My first stop would be to see what the big humanitarian agencies/providers already collect to monitor needs. I attached this one as an example, but organizations like UNICEF, Save the Children, Doctors without Borders, maybe the World Bank, all likely maintain big independent data bases that should give you a starting point.
    • Aug 31 2012: Thank you Fritzie. I will definitely check that out.
  • thumb
    Aug 31 2012: It all started in Jacksonville Florida with a friend of mine. He wrote a program that tracked the sale of rice in a certain, large grocery outlet. He was able to establish a graph that predicted the highest probability of customers having used the rice at home and coming back to buy more. When the time came, the store led off with ads to imply sales but the price rose just as the greatest number of customers were arriving to buy the rice.

    They also discovered that people brought rice in larger bags because the price per ounce was lower. That changed (before they started posting labels telling you the price per oz.) to increasing the price of large bags while it was actually cheaper to buy the smaller bags.

    What surprised me in all of this is that people took years before they started paying attention to the real price of food per oz.

    No matter what model you create, it will be used to exploit the customer. The goal of all business is to get people to buy more than they need at a higher price. That's why, today, you pay almost $3.00 for a loaf of bread that cost me in 1980 only 56 cents because they can use analytical models to understand when and why people go shopping to maximize their profits.

    It cost 10 cents to make a can of coke plus 2.5 cents for the can. With shipping and handling it cost 19 cents to put it on the store self. It has zero nutritional value.