Published on the 08/09/2010 | Written by Newsdesk

In the old days, mapping data was easy. You bought a wall map and lots of colourcoded pins. Then you stuck them on the map. Fast, easy and cheap. But the information content was superficial at best…

Then came GIS (geographic information systems), it was slow, difficult and expensive, but you could get amazing results. These included brilliant analyses, beautiful cartography and a whole new insight into your data.

However to get all this, costly digital base-maps needed to be developed; you had to prepare attribute data and buy complex software and powerful workstations, to process all the points and polygons. You also needed a team of geographers in the back room who could also program.

Then came Google Maps and Digital Earth which offered incredibly detailed online base-maps and GPS (global positioning systems) for capturing location coordinates. PCs also became more powerful and mainstream databases could now store and process spatial data. A whole new world has opened up as a result: mapping data has become faster, cheaper and easier. The challenge now is to ensure you map it right.

It’s rather like driving a high-performance sports car: you have tremendous power to go fast, but one small lapse of attention could send you over a cliff.

Location intelligence is a powerful tool that can significantly enhance decision-making capabilities, but it can also trick users into making bad decisions if maps and data with different levels of accuracy are combined.

This question of accuracy is a major focus of research in GIS science and it is no less important in location intelligence.

“From an implementation point of view,” says John Hoffman, CEO of Altis Consulting, an Australian-owned consultancy offering specialist expertise in data-warehousing, business intelligence and data-management, “it is significantly easier to provide ‘location intelligence’ now because BI vendors provide map integration capabilities in the new versions
of their software.

“However, the big issue is still the quality of the information, especially when it comes to GIS coordinates. End-users are excited about the possibility of viewing information and detecting patterns and relationships, via geographic or building maps. But they’re often disappointed because the location data quality is poor. BI vendors have done their part, now we must improve the quality and structure of the data to provide consistent and trustworthy information and insights.”

“The real need is to get the back-end right” says Simon Jellie, managing director of Wellington-based e-Spatial.

“If you get the back-end data correct – accurate location data for your customers or events and a map base that is the appropriate scale and accuracy to support that data– then the front-end will provide a powerful insight into your business,” says Jellie, who is a 25-year veteran of the digital spatial data market.

“If you are going to implement a spatial location solution for your business intelligence data, you have to ensure that your data is clean from the outset. This is our specialty– ensuring that the data is correct before it goes onto the maps. It’s the same old story: garbage in, garbage out.

Scrubbing data is a time- and labour-intensive process, but absolutely necessary.”

So, it’s all about the data. In columns and row, a few bad metrics are almost invisible, but any BI user will tell you that as soon as you start showing data in graphs or other visuals the bad data stands out. And, when you map, the errors are magnified. If you are going to add location intelligence to your BI capabilities, start with clean data. Otherwise, the maps you get could steer you in the wrong direction.

For more information visit the iStart Business Intelligence Pavilion