For Business Intelligence, the Trend Is Location, Location, Location
Sixto Ortiz Jr.
Businesses are always looking for better ways to analyze the large amounts of data they receive from customers, suppliers, market analysts, and other sources.
Modern business intelligence approaches began in the late 1980s, but companies have used of some type of BI since business-support systems were introduced in the 1960s.
Now, BI is beginning to include a critical new element: location. Location intelligence is the ability to organize and understand complex events and trends by studying the geographic relationships in information.
In essence, LI adds geographic, demographic, economic, and similar types of data to the financial, marketing, and other information already used in BI.
This reflects the growing development and use of affordable technology that can capture and recognize geospatial data.
However, LI still faces several key challenges.
Early adopters of computerized LI in the late 1980s included telecommunications, oil, and gas companies, as well as government agencies, utilities, and others for which location-based data was critical. However, the cost and complexity of adding geospatial data and functionality to existing BI applications limited its use.
Technology advances have recently enabled the growing availability and affordability of geographic services such as Google Maps and Microsoft Virtual Earth.
Not long ago, only geographic information system (GIS) experts used such services. Now, though, the development of turnkey applications that automatically process tabular data into information that can be plotted on maps, coupled with sophisticated analytics software that can process such data, have brought LI within reach of many organizations.
At its most basic, LI entails the assignment of location data and other relevant information to a database, explained Mark Smith, CEO and chief research officer at advisory services firm Ventana Research.
Accomplishing this manually could require considerable programming.
To automate the analysis of multidimensional geographic and nongeographic spatial information within a single database, LI uses technologies such as spatial online analytical processing.
In evaluating the data, LI uses standard BI tools such as predictive analytics and business-process intelligence.
After analysis, LI plots the results on a map, which requires geocoding and the extraction of geographic coordinates from textual address data.
Applying geocoding to large amounts of data requires customized programming that is beyond spatial databases' basic functionality. This programming can be difficult, time-consuming, and costly, explained Brandon Purcell, vice president of engineering at LI vendor SpatialKey.
However, LI applications include software that automates the process of translating textual address information into geographical coordinates that can then be plotted on a map.
Some companies are now using LI to locate new businesses such as gas stations and convenience stores, noted Jim Harder, a principal with the Visual Data Group, a BI vendor.
The technology lets companies analyze the local demographic, economic, and other relevant business-related information to determine whether placing a store at a particular site makes sense.
Businesses can also use LI for purposes such as market-penetration studies.
Cellular-phone companies use the technology for network planning and locating cell towers. Government agencies utilize it for urban planning. Insurance companies employ the approach for risk management. And real-estate agencies use it for site reports.
Even law-enforcement agencies are working with LI to analyze crime-related data and the geographic distribution of incidents to identify "hot spots" where they should concentrate their resources, said SpatialKey's Purcell.
The approach could also be used with augmented reality to provide, for example, the real-time mapping of complex location information over a display of a user's surroundings.
LI faces several obstacles to success.
For example, noted Harder, keeping mapping information current with the addition of new streets, housing, businesses, traffic patterns, and other important factors can be challenging.
In addition, many companies may not understand LI's business value or how to best use the technology, said Purcell.
Businesses also sometimes have trouble integrating the different software they need to load, report, map, model, and process LI data along with their traditional information, added Ned Harding, chief technology officer at LI vendor Alteryx.
Down the Road
The continued integration of spatial information with other types of data could make LI even more useful, said Harding.
As users want to process increasing amounts of location-based data, he added, they will move to grid-computing approaches using large clusters, which will require a bigger IT infrastructure.
By eliminating the need for users to house huge analytical applications on their systems, cloud-based approaches will make it easier to collect, distribute, and study LI data on a large scale, said SpatialKey's Purcell. This will make the technology available even to small and medium-sized businesses, he added.
Also, said the Visual Data Group's Harder, LI tools will grow in sophistication and will feature expanded data storage and management coupled with fast and easy information access.
A key question is whether businesses will embrace LI as yet another way to study data among an already dizzying array of business-analysis tools.