Building Bridges with Location Intelligence

Next-gen Business Intelligence
By Victor Harrison | Published July 26, 2010

Businesses are at various levels of maturity relative to investments in business intelligence data and systems. For the most part, Business Intelligence (BI) as an IT and data investment has been the domain of Fortune 500 companies. Advances in cloud computing and SaaS/managed services bring BI to small and mid-sized companies and offer compelling solutions to business users of large organizations that struggle with getting the information they need and want from their data and IT. In particular, what is increasingly becoming the norm is small and mid-sized companies that have big company data issues. (For an example, see sidebar on Stubb’s Bar-B-Que.)

Location intelligence (LI) is the bridge between business intelligence and market intelligence, the bridge between the database analyst and the business user, and the bridge between the integration and correlation of various enterprise data systems such as CRM (Customer Relations Management), ERP (Enterprise Resource Planning), and BI. Location data has been the missing link in systems that allow companies to optimize and leverage their investments in data and to recognize the promises made by vendors that data analytics offer, in fact, a competitive advantage. BI is an analytical system that usually relies on data from transactional operational systems such as CRM and ERP. Figure 1 illustrates the conversion required from transactional systems to analytical systems.

Location intelligence is more than just the ability to plot data on a map, although visualization is a big component. LI is the next evolution of business intelligence, because:

  • It provides BI with a value proposition beyond large companies and makes BI available to mid-market companies;
  • It offers a cost-effective solution to complex data analytics;
  • It extends the existing BI infrastructure and environment of companies that have already invested billions of dollars in these systems by making BI solutions scalable across the enterprise and down the food chain; and
  • It brings the data closer to the user and makes it more actionable.

We spoke with four senior LI company executives from the field to gain their perspective on this subject: Tom Link, Owner/CTO, SpatialKey; Jim Pollock, President of Vertical Markets, aWhere; Jean-Sebastien Turcotte, Executive Vice President/CTO, KOREM, Inc.; and Luc Vaillancourt, CEO, Spatialytics. Their comments are included in this article. These companies provide consulting, strategic and integration services, as well as next-generation LI tools, technologies and solutions with a focus on keeping LI simple for the customer.

LI and BI:The Differences, Advantages, and Options

All of those interviewed unanimously agreed that business data are already spatial because all businesses are place-based entities, and that LI is a subset of BI. Databases such as Oracle Spatial, Teradata 13, and Neteeza Spatial have now all made the storage and consolidation of location information across business data easily accessible. The main difference between LI and BI from a user perspective is the difference between the consumers and producers of data. Tom Link of SpatialKey explained, “the trend in BI has been to focus on the production of the data—the data warehouse, the database analyst, the cube, which often results in rigid reports that the business user has to figure out how to apply to day-to-day management. LI, on the other hand, is about the trend in visualization, usability, interactivity, and shareability of the information. The focus of LI is really on the person that has to consume and use the data to make decisions.”

Many people confuse the map with location intelligence. The map is definitely in many cases a better way of visualizing data that is already in a BI system. See Figure 2. for the pattern recognition available from displaying tabular data on a map, which is unrivaled when geographic distribution of the data is in fact the insight. However, there is much more to LI than just plotting data on a map. Figure 3 illustrates the data sources, analytics and technical platform required to deliver location intelligence.

It’s All About the Correlation

The power of location intelligence is in the ability to correlate data down to, for instance, the location of a store, a customer, a medical clinic, etc., in relation to location influences such as weather, culture, socio-economics, or competition. “The real location intelligence is making use of great things that come out of GIS (geographic information systems) and bringing them into a BI workflow and being able to ask very sophisticated questions, such as, ‘Where are my stores that are underperforming relative to competitors’ stores?’ as opposed to the simple sales question, ‘Where are my stores performing or underperforming?’” said Tom Link of SpatialKey.

Jim Pollock adds, “Often, we discover hard-to-visualize similarities through mathematical spatial analysis. Conversely, we often see visual patterns in a map that are hard to do mathematically. Maps and numbers are great complements, not either/ors. Mapping is part of this, but just one more visual input. It helps to visualize patterns that emerge or do not emerge. If no pattern emerges, that could be a message, as well.”

What Are the Options?

Traditional BI analytics are limited because they generally don’t consider location. According to Jean-Sebastien Turcotte of KOREM, “The geographic dimension is often marginalized because traditional IT rarely assesses the value of geo at the beginning of the project.”

Companies that wish to own their own system, therefore, have two options when it comes to LI, according to Luc Vaillancourt. “For new projects, they can develop a fully integrated GeoBI system that incorporates the geographic/location dimension from the beginning, or for existing legacy systems they can retrofit the system with a geospatial visualization component, and LI acts as a bridge to these existing data systems.”

A third option is becoming increasingly important to small and mid-size companies with big data issues and small IT budgets, and for executives in big companies with big IT budgets who are frustrated with the quality of the information they can extract from their systems. The third option is a managed service that is offered by companies like aWhere and SpatialKey. With a managed service, a company can take advantage of the spatial/location analytics of the provider’s location intelligence platform, and the company doesn’t have to worry about the most important and expensive component of a location intelligence system, the data—the external third party data, including the mapping data.

According to Turcotte, “Businesses always underestimate the costs associated with the data, which are typically 35-45% of the cost of an LI project. Also, 30-35% of the initial data investment needs to be factored for annual maintenance of the data. When we implement such projects, we can spend 20-50% of the time structuring the data to make it insightful, and when we go back a year later, we find that the company hasn’t been maintaining the data; the data maintenance is the most critical component and is where most companies cut costs. Accuracy of the data is where the data costs rise; the more dynamic the data, the more often you have to update the information.”

In today’s highly competitive global marketplace, which is in a continuous state of transformation, possessing the ability to visualize and analyze BI quickly and affordably becomes a critical key to success. Now, companies of all types and sizes can leverage sophisticated LI in a predictive mode for planning and strategic purposes and can gain a better understanding of trends and patterns from their data as well. For example, an insurance company can leverage third-party ‘cause-related’ data to help predict severe weather patterns and conditions such as hail storms. They can then notify their ‘at risk’ customers in advance so they can adequately protect their family, vehicles, homes and other assets. This LI can benefit all by increasing safety and reducing claims, while improving customer service and enhancing the overall customer experience.

“BI is an analytical way to look at business. Spatial BI is taking into account the fact that your business is affected by local variables such as weather, demographics, events, and topography. New LI tools expand this. LI is a component that often needs to be a part of your BI superset. The ratio depends upon the type of business you are in.”

Jim Pollock, aWhere

The good news for businesses today is that they create and have access to enormous amounts of stored and real-time data-based information or business intelligence (BI). In the past, it would often take large investments in technology and talent to effectively mine, convert, analyze and leverage one’s BI. Today, location intelligence (LI), the ‘seeing is believing’ aspect of BI, can help to bridge this challenge by providing companies of all types and sizes with a new generation of tools and technologies, along with innovative business models and open platforms to affordably visualize and analyze their data.

“ Businesses tend to buy lots of third-party data. It usually sits on a shelf because they don’t know how to use that data. For example, a medical company buys demographic data to help to identify which hospitals to approach. What they get is a file that someone looks at for awhile but doesn’t know how to integrate. LI is often a way to correlate different sets of data.”

Tom Link, SpatialKey

Is your data collecting dust on a shelf, and tucked away in a not-so-friendly knowledge management system? Managers can now fully take advantage of the investments in third party data. It’s time to get closer to the data, to understand all its benefits and revelations. Data and systems that are easy to use, will be used.

“Whenever your problem is related to distance of distributors, markets or distribution of data, then you need to spatialize the data in order to discover the relationships. As soon as you exploit location information, you are doing LI; BI that is location aware is LI.”

Jean-Sebastien Turcotte, KOREM

Time and space are key correlating factors. Time has been managed relatively well in BI systems and not well in GIS systems. Space, on the other hand, has been well managed in GIS systems, but not BI systems. LI is the bridge. Today, the development and availability of scalable and affordable LI tools, technology, talent and services enable and expand access, allowing the small and medium business community to also leverage their data-based BI. These companies now have access to information a GIS department would ordinarily provide to enterprises, extending the BI value proposition that in the past was mostly the domain of large business.

“BI is about day-to-day decision making and strategic planning. This means that large amounts of apparently meaningless pieces of data have to be restructured into multi-dimensional cubes to make sense of the data—to identify the crucial information needed by management. LI is the bridge between geo and existing systems like ERP, CRM and e-commerce. With Geo-BI, it’s more than a bridge; we fuse BI and Geo into one Spatial OLAP (Online Analytical Processing) system. This makes it easier to feed, to manage, and to scale. And, more importantly, it allows quicker, more direct and powerful analysis and visualization capabilities with maps and spatial queries, in sync with charts, graphs and other visual gimmicks we usually see in BI dashboards and reports.”

Luc Vaillancourt, Spatialytics

Some of us can look at both raw and refined data and see patterns and trends that, to most of us, are invisible or unrecognizable. It is rare to find a marketing, sales, product management, strategy or C-Level executive who has the power, skills and ability to do this. As a result, companies could be missing types of signals that can be relevant to their business, directly impacting the bottom line, particularly if they are not able to look at correlations and causes for growth that predict upward and downward demand. This problem can be complicated by numerous disparate data sets, with no cost-effective linking mechanism or available integration capabilities.