Category Archives: Open Data

Connected: Jack Dangermond, president, Esri

I feel very conflicted about this article.  On the one hand it’s great to see a mass business media publication like Forbes writing about the importance of GIS to business, on the other hand, I am saddened to see that Jack, and the article refer to the business applications as “emerging”.  I’d argue they are not emerging, but have been there.  I have spent the better part of my career establishing the use go geographic or spatial analysis in the business context, and some very large companies are known for deploying GIS.  ESRI has, admittedly been late to that game (MapInfo – now part of Pitney Bowes – was the dominant player for a while), but the work the truly disruptive force in the space are MapBlast (acquired by MSFT), Google, and MapBox.  They all share in common the fact that they truly democratized the use of maps in business by a) injecting new ideas/approaches, b) becoming active contributors to an open (source/data) movement, and c) embracing a diverse ecosystem of players.

Connected: Jack Dangermond, president, Esri.

2014-06A: From Around The Web: The backlash against big data, continued – O’Reilly Radar

A good articulation of the issue with “big data”.  Data is here (and has been for a while), but we have to be careful about the false assumption of what more data can do.  One thing that Mike tangentially touches on, but is worth highlighting is that we have to continue to be careful to let data drive all decisions.  We need a bit of a return to a balance of qualitative and quantitative, not a continued push towards quantitative and no qualitative.

The backlash against big data, continued – O’Reilly Radar.

2013-08: Demographics of 1

A very quick post to get some thoughts down on paper.

I used to work in consumer marketing in banking.  One of the regular habits was trying to understand our customer base and our prospects.  For that, we relied heavily in segmentation.  Segmentation, is nothing more than trying to take a group of people (say all your customers) and try to group them or organize them into sub groups that allow you to understand how they behave or what their preferences may be.  In banking, we often use life stage market segmentation, as you can rationalize people’s financial decisions and the opportunities to market to them through the stages of life that they go through.
For example, one could argue that many people go to college, and to go to college they will need to borrow money.  This is a great time to talk to them about student loans.  Therefore you could train your marketing segmentation to find all people in the age of 16-21 and who either are in college towns today, or live in areas that have college degrees (as there’s an assumed correlation between presence of degrees and likelihood to attend.
One can keep on coming with similar examples of why it makes sense to segment the population into logical sub-groupings, and trust me, the nefarious applications of this are not lost on me (google red-lining).
But this is a very antiquated model.  It’s a model that’s based on a) a lack of availability (either due to internal technical means, or simply lack of capture) of data on customers, b) antiquated concepts of human behavior, c) focused on likely models (i.e. normal distribution of data, “we have a high degree of confidence that this large group behaves the same”), and d) not appreciative of the ways in mobile and digital technology have vastly changed how we behave.
So the next wave is the marketing to the individual… Marketing to 1, segmenting to 1 or demographics of 1.  What does this mean?  Basically that we need to develop mechanisms for marketing to the individual at the right point in time or place.  Therefore making your (the marketer’s) connection to the individual extremely relevant and pertinent to the moment and/or the place that you are in.  For example, a bank should start to think about prompting me with options for banking when it studies my behavior and sees when I (not the people in my age, income, geographic segment) use their services, and predict what I will I as an individual will need. It’s that magical moment of relevance that is so well captured in movies when the relationship between two protagonists is epitomized by one knowing the right details about the other, without that other having described them.
We are starting to see this develop with push towards geofencing, mobile advertisement, etc…  But how do organizations get themselves ready to develop marketing strategy based on this new data landscape, and more importantly how do they shift from their current antiquated models and data systems to these newer systems.