More and more businesses today compete on the availability and accessibility of their information. Fortune 1000 companies tapped into business intelligence technology many years ago, and today these large organizations rely on the information provided by this technology. In recent years, business intelligence technology has started moving down-market. The number of software packages serving manufacturers that offer business intelligence modules is increasing. But, given the choice, many manufacturers are still opting for simple reporting solutions instead of investing in business intelligence.
Why? Because they are not aware of the depth of information they can access with a business intelligence tool. By understanding the power of business intelligence, manufacturers can begin to change the way they think about business questions. In many cases, asking a "smarter" question will give manufacturers a competitive edge in the marketplace.
Before exploring business intelligence technology in greater depth, it is important to understand that a manufacturer must have a host system in place that can capture the right information--an integrated system that collects accounting, distribution, and manufacturing information--not different, unconnected systems that require duplication of information. The host system and its databases form the foundation for business intelligence technology.
Once that solid foundation exists, the next step is to evaluate the business problem, and then find the technology suited to solve that problem. To do this, manufacturers must determine how much data is needed to solve the business problem, how often the needed data changes, whether real-time access to information is necessary, and who will be accessing the data.
Inventory
Instead of asking: How much inventory do I have?
The smarter questions are: Where did my inventory came from? What was it used for? Who was it sold to? The problem with the first question is that its answer doesn`t provide a piece of information on which a business can act. Instead, it is just a simple data point.
The smarter questions center on the sources and uses of a company`s inventory. The answers will help a manufacturer determine whether inventory came from an outside vendor or from its own manufacturing floor, how old the inventory is, and how fast inventory turnover occurs. Knowing what the inventory was used for--as part of another product or sold directly to a customer--is also valuable information.
Putting the two sides of the equation together is possible, especially when manufactured items and their components are specifically identified by serial numbers or when one lot number is used to identify a production run or batch. Once this information exists on a manufacturer`s host system, business intelligence tools spring into action.
By understanding the power of business intelligence, manufacturers can begin to change the way they think about business questions. In many cases, asking a "smarter" question will give manufacturers a competitive edge in the marketplace.
For example, serial lot tracking is crucial (in fact, mandated by the government) in the pharmaceutical industry. Production is expressed in terms of lots, where one product is produced in different lots over the course of several weeks. Lot numbers and expiration dates are stamped on products. In the event of a product recall, an entire lot might be recalled in the interest of consumer safety.
Using data mines, one form of business intelligence technology (see sidebar), the pharmaceutical manufacturer has the ability to identify which ingredients went into a particular production lot and where that lot was shipped. If a customer calls the company to complain about a "suspect" product, an employee can easily enter a product or lot number into the manufacturer`s data mine and access all other customer orders where that lot number was sold--then notify those customers of a recall, if necessary.
Purchasing
Instead of asking: What do I need to buy?
The smarter questions are: What materials will be used to make product A for the next month? Do I have enough material in stock? Do I have the materials on order with vendors, or do I need to order more? Will the materials get here in time?
No matter how well a manufacturer plans production schedules, material requirements and purchasing activities, the first question inevitably turns into "What do I need to buy today?" at the start of each business day.
One important thing to remember is that purchasing is not a distribution, but a manufacturing task. Before a manufacturer explores business intelligence technology options, it must have an effective manufacturing planning and control system--a data warehouse that integrates production scheduling, materials planning, and purchasing. This data warehouse needs to be rebuilt from time to time so that planning is realistic and control is possible. Once the data warehouse is built, ad-hoc queries, data mines, and reports can be used to access needed information.
For example, a manufacturer of computer components may use a data mine to determine what materials need to be expedited or purchased today to fulfill a particular customer order that will be due in two weeks. In addition, this data mine can compare the number of finished goods to current inventory levels, the forecasted sales to actual sales over a number of months, and the actual purchase orders vs. the planned ones for each component item. Data mines can help the manufacturer understand the multiple levels of a product as well.
Sales & Production
Instead of asking: Are sales strong?
The smarter questions are: What are total sales compared to shipped orders and paid orders? What is the composition of sales by product, product category, territory, or salesperson? How do sales compare to last week, month or year?
Although looking at sales levels at a given moment is a good thumbnail indicator of a manufacturer`s success, drilling down to the details gives a business the information to make smart business decisions that will help ensure a company`s longer-term success. Examining information such as shipped and paid orders helps a company pinpoint any manufacturing or collections issues it might face. Determining the composition of sales by salesperson can help a manufacturer reward top salespeople while discussing potential problems with those who are not meeting their goals. Business intelligence lets manufacturers get to the heart of sales information and make appropriate business decisions.
The more complex sales-related questions can be best answered by constructing a data warehouse. Each subcategory sales record--for product, salesperson, or territory--is placed in a "bucket" for a particular time period. Because the data warehouse is off-line, analyzing sales information does not interfere with daily processing and can be performed at any time.
Once the warehouse is built, employees can use ad-hoc queries to analyze each sub-category or a combination of them for specified time periods. For more extensive analysis, data mines can be used to drill down from summary sales information to the details that make up the summary amounts. Sales trends can also be used to improve sales forecasting, which leads to more accurate production scheduling.
Production & Manufacturing
Instead of asking: Am I producing too much? Am I producing enough?
The smarter questions are: How do customer orders compare to actual production? If a customer order is directly tied to a production order, has the product been built? If an item is backordered, when can I promise delivery?
Some host systems allow customer orders to be tied to production orders in a quick-turn environment. That is, production is based directly on sales. Such a host system can also help a manufacturer keep customers informed as to the current status of their orders, providing valuable information such as when the order will be built, and when it will be shipped.
For more extensive analysis, data mines can be used to drill down from summary sales information to the details that make up the summary amounts. Sales trends can also be used to improve sales forecasting, which leads to more accurate production scheduling.
In this type of manufacturing environment, where production is directly tied to customer orders, a data mine can be used to drill down from the customer order system to the production order system, determining the current status of the production order, including quantity completed, planned completion date, and current availability of component items that are required to make the product. If the production order is dependent on the production of sub-assemblies prior to final assembly, data mines can be used to drill down one or more levels of the product structure to determine the current status of production orders for sub-assemblies. In more traditional production environments, production is based on pre-determined schedules rather than actual sales. Manufacturers must use information from their production schedules, inventory levels, and current product commitments to determine when new orders can be completed and shipped. Here, a pre-defined data warehouse can be developed, and SQL commands or queries can be used to populate the structure. The data warehouse helps a manufacturer determine availability levels for a product over a period of time, taking into account current supply from inventory, available or forecasted supply from purchase orders, data from production planning and production execution systems, and committed demand from customer orders.
Understanding the depth of information business intelligence systems can provide is the key to making an informed decision about investing in this technology. Business intelligence provides manufacturers with a detailed level of information and analysis that can help companies make better business decisions. Asking "smarter" questions and using this technology to find the right answers can give manufacturers a competitive edge in the marketplace.
John Antjas is vice president of systems technology, and Wendell Giedeman is director of applications design and documentation for Macola Software, Marion, Ohio.
Business intelligence technology defined
Selecting the right business intelligence technology must be done based on an accurate definition of a business problem. Once the business problem is well understood, manufacturers have several business intelligence technologies to choose from:
Query technology
The interface for query technology is graphical, allowing a novice user to quickly learn the technology and extract needed data from a manufacturing system. Often, queries work best in accessing information that is stored in one or two tables with simple relationships. Because queries are easy to build, they can be constructed on an as-needed basis. Some queries may be useful throughout an organization and can be added to desktops for easy access.
Report technology
In general, reporting is defined as the extraction of information from the host system database and the transfer of that information to some media that can be disseminated throughout an organization. Strong reporting systems have a powerful language embedded, allowing users to develop reports of differing complexities. Some reporting technology can collect data from different systems, allowing users to analyze consolidated information across the enterprise. Developing complex reports should be done by staff who have a strong understanding of the company`s database and its relationships to other systems. Once developed, reports can be provided to employees for daily use.
Data mine technology
Data mines are graphical representations of data relationships. Simple data mines can be represented by spreadsheets, while complex database relationships must be represented by an entity diagram. Sophisticated data mining technology allows employees to extract data from extremely large, complex databases in real time. Data mines answer complex questions and can link different systems together. For example, an executive could review the inventory levels at three different manufacturing plants or an employee could analyze data from both a current system and an older legacy system. Data mines solve more complex problems faster than queries.
Data warehouse technology
Data warehouses, sometimes known as data marts, can be used to simplify complex databases, allowing more novice users to meet their own business intelligence needs. A data warehouse is an assembly of information from various sources, which is pulled into one, centralized database or set of tables. Using this technique, information from several large, complex databases can be converted to a more simple database, which is used solely for business intelligence purposes. The data in the warehouse is not synchronized with the live data at all times, but the process of building a data warehouse can be automated, allowing regular updating of information. Once collected, information can be extracted, messaged, and analyzed using ad-hoc queries, data mines and reports.
Share
