The Intelligence Triangle

Strategic Security
The Intelligence Triangle
 

​The fire triangle is an illustration commonly used to describe the elements required to make a fire: fuel, heat, and oxygen (or an oxidizing agent). If any one of these three elements is missing, a fire cannot occur, and removing one of these elements from a burning fire will extinguish it. Applying this metaphor to an in-house intelligence function reveals three critical sides to framing the intelligence program: a strong information requirement, access to a variety of data, and the tools and expertise to process that data. Here’s how one national retailer used this intelligence triangle to deliver a comprehensive business plan and technology integration.

Information Requirements

Information requirements make up the bottom of the intelligence triangle, analogous to fuel. Information requirements typically come in one of three types—standing requirements, ad hoc requests, or information discovery projects.

Standing requirements might include regular reviews of financial reports, a weekly summary of industry news, or supply chain analyses. These reports should provide data on the general state of day-to-day business and should come in a standard, easily digestible format. The most efficient intelligence programs find creative ways to reduce the manpower burden of standard reporting through automation, standing data queries, and simplified production.

Ad hoc information requirements result from an anomaly in regularly occurring reports or from an information gap in current standing requirements. A competitive analysis of a rival organization or open-source research on a prospective client are examples of ad hoc needs. Ad hoc information requirements will call for custom reports and data sources.

Information discovery projects use creative analytical processes to test new data sources or to seek out lines of thought from existing data that may not be apparent. A common goal for an information discovery project is to identify links between seemingly unrelated data sets.

Information requirements can vary among internal departments, so they must be prioritized within the department and across the enterprise. Information expires over time and must be regularly reviewed, especially with standing requirements. Because standing requirements become part of the day-to-day operations, a monthly or quarterly review of information requirements will ensure that resources are allocated to producing only relevant reports.

The retailer, a client of the author, initially focused on information requirements, assessing data that was currently available, and identifying information gaps. The retailer’s long-term objective was to expand its footprint by 250 percent over the following five years. The first phase of this project was a complete and current description of where the organization stood with some initial hypotheses on successes and shortcomings backed by readily available data.

During the information requirements portion of the project, the company analyzed internal data—transactional information, foot traffic, and sales trends. These metrics also formed the foundation for the retailer’s standing information requirements. These were daily, weekly, and monthly metrics required to make regularly occurring business decisions on issues such as employee scheduling, supply chain and inventory management, and marketing promotions. The analytics were primarily descriptive in nature, but began to give a sense of where the company might dig deeper.

Access to a Variety of Data

Data variety and access is the side of the intelligence triangle analogous to oxygen. Data sustains intelligence by revealing hidden relationships in information. Understanding what data is available right now and what data is missing is a critical step. Once data gaps are identified, resources must be applied to acquire new data.

Companies should consider broad ranges of data types, such as operational information, sales, financial, current events, government policies, cultural information, social media, regional demographics, access control data, and weather as starting points for data collection. Companies should encourage data sharing across different departments.

Phase two of the retail project took a step beyond descriptive analytics and started to identify data sources that could fill the client’s information gaps. During the initial phase, a number of intelligence gaps were identified, such as a limited understanding of demographics at county level or below for each location and inadequate data on current traffic patterns, competitor locations, and local government development plans across the organization’s footprint.

Additionally, some external events were in sync with irregular business patterns, like a substantial increase or decrease in sales. Analysts considered factors such as weather, local entertainment events and sports competitions, nearby business openings and closings, and marketing efforts. This step allowed analysts to describe the external factors that coincided with favorable business history.

These projects splintered into ad hoc information requirements like a 90-day traffic pattern analysis and sales number studies referenced against weather events. The next step was to take on discovery tasks, such as a competitive analysis of other stores in the area and the scouting of future locations for long-term expansion.

Cutting-Edge Tools

The last side of the intelligence triangle is communicating the analysis to a broad audience, providing knowledge, which is analagous to a fire giving off heat. Often, analysts will spend days or weeks deconstructing data to reveal new knowledge, but will only have 15 minutes to communicate that knowledge to the intended audience. Analytical and data visualization tools enable reporting that is easily understood by clients and executive decision makers.

Structured data is easy to analyze since data is broken out by predetermined data fields. Typically, this is the type of data that fits into a spreadsheet or form. Names, addresses, financial data, and other categorical information can easily be manipulated in tables, charts, and graphs. This represents basic descriptive analytics. Descriptive analytics count and display data, which is critical for understanding, but they do not really talk about the underlying meaning behind the data. Structured data can be measured across a number of different data types, but as the data sets increase in dimension, the complexity of data processing, analysis, and visualization goes up. That is a good thing, so long as adequate tools are used to communicate the analysis.

Unstructured data is a bit messier to work with. This is the data that comes in the form of free text or nontabular information. The analyst has to identify what information is of value, mark it, analyze the aggregate data, and draw a conclusion. Many business intelligence software solutions specialize in modeling unstructured data, creating libraries of key words, and quickly blending the data to produce reports. Expertise in business intelligence software is required when tailoring models and reports for custom or ad hoc information requirements.

Phase three of the retail project shifted focus to presenting the information to executives. The information would be used for long-term planning, such as selecting locations to open new stores during the organization’s expansion. Analysts gained a strong sense of what made current locations successful from the data. The task ahead was to discover locations that matched similar criteria. The research allowed the retailer to secure additional financing and the new stores are currently under construction.

Establishing an in-house intelligence function creates measurable value when informing key decisions. Harnessing the elements of a construct like the intelligence triangle enables decision makers to be confident as they navigate through the market. 

Bryan Lane is senior associate at global RISK advisors, specializing in data analytics, assessments, and strategic planning.