Business Intelligence as a Strategic Weapon


Business intelligence (BI) has come a long way in recent years. In the wake of Big Data and information overload, enhanced search methodologies and advanced systems have surfaced that fine-tune business intelligence to swiftly deliver targeted pieces of knowledge that can inform decisions and drive actions. Yet many organizations underutilize BI, which can have a significant impact on profitability.

This article describes the role new business intelligence can play, what capabilities it must have to effectively impact profitability, and a few examples of how producers, suppliers, and investors in the paper industry can and should deploy it as a strategic weapon in today’s competitive landscape.

No one makes a profit simply by owning papermaking assets. A business’s assets—people, plants, and products—while different, are not unique. Anyone with access to enough cash or credit can buy or build a pulp or paper mill, hire people to run it, and make products using standard recipes. Take Eldorado, for example, a 1.7 million metric ton bleached eucalyptus pulp (BEK) producer in Três Lagoas, Brazil. The company, which started producing five years ago, was owned by a family with capital but no prior experience in pulp and paper. (Norwegian firm Paper Excellence recently purchased Eldorado in a deal worth R15 billion, about US$4.7 billion.)

Sustainable profit comes from the value gained through constant refinement of utilized assets. In commodity businesses like pulp and paper, that effectiveness is driven by a wide range of decisions which, when made marginally better, result in higher prices, higher operating rates, lower costs, and lower volatility than in less well-run firms. So, while capital investment is the cost of entry, profitability comes from superior decision making. That requires two things: the skill of the organization’s people; and relevant, timely, accurate information that informs their decisions.

The same is true for the industry’s suppliers. If there is one challenge most suppliers face, it is commoditization, where price becomes the key determinant of competitive success and customers require increasing amounts of technical and sales service without being willing to pay a premium for it. The antidote to commoditization is differentiation, where one supplier demonstrates an advantage over competitors that is valuable enough to warrant a premium. Finding and demonstrating valuable differentiation requires suppliers to understand customers’ unique needs deeply and precisely, which is a role business intelligence should support.

Traditionally, professionals in the pulp and paper industry turned to personal experience and directories for information. In recent years, however, pulp and paper has globalized to the point that no one person can get first-hand exposure to all relevant parts of the industry. Companies themselves are bigger and managements’ staffs are smaller. Experience and intuition, while vital, are only a piece of the knowledge base that is required to guide today’s firms. Directories, the other traditional information source, have severe analytical limitations and data quality shortcomings (more on that to come).

Owners and managers today need a new breed of BI with a new set of capabilities—one that can both aggregate massive amounts of data and refine it to address the specific goals at hand. To deliver that, business intelligence resources need to have fundamentally different data quality and functionality characteristics. But what does it take for business intelligence to perform at this new level?

Just having information doesn’t create profitability. The key is how the information is used. We at Fisher International call this “data-driven decision making.” Thirty years of experience investing in the resources required to support data-driven decision making in the pulp and paper industry, and coaching in their use, has taught us the principles that make business intelligence work.

One of the most important is that senior management must instill an expectation of driving decision making with data and analytics. Data-driven decision making is an organizational skill that differentiates a company from its peers. Skills and habits develop only with practice, so senior management’s insistence is a key success factor. This is Principle #1: Getting full value from business intelligence requires a corporate culture of data-driven decision making.

Decisions people make in the paper industry are expensive because of the industry’s scale and capital intensity. So, business intelligence and data-driven decision making deliver a lot of leverage—which can be positive or negative. This means that BI support must be highly reliable. Just as someone would not build a critical product like an airplane without rigorously testing the materials, one should not base a company’s strategy on business intelligence without assurance of its reliability. Thus, Principle #2: A business intelligence system must be quantifiably reliable; this can happen only by meeting certain requirements, one of which is near-perfect data quality.

Another important factor is that each business decision is unique. Every company’s situation is at least a little different from its competitors’, and those circumstances are constantly changing with time. Business intelligence support must be both flexible and precise—it must answer any question business people can express and it must answer that question exactly, at that precise moment, for that particular need. Traditional sources like look-up directories, canned reports, newsletters, multi-client studies, and other commodity sources of news and information don’t do that. This is Principle # 3: A business intelligence system must have the power and flexibility to address any differentiated, precise question its user asks. This can happen only with enough data detail, foundational database structure, and tool design.

While the cumulative impact of companies’ business decisions is huge, the cost of business intelligence that informs those decisions is actually extremely small. Its low cost gives business intelligence users incredible leverage, making investment in business intelligence one of best returns available in today’s environment. The critical impact of data-driven decision making also means that one would always prefer to have more reliable, more capable business intelligence support if it were available. Principle # 4: It is almost always worth investing in more capable and reliable business intelligence.

Almost everyone uses business intelligence to some extent. Over the years, Fisher has observed pulp and paper professionals’ use of information and decision-making processes and we’ve identified the key functions business intelligence should provide. We’ve segmented those best practices for each type of company in the pulp and paper industry (producers, suppliers, investors, etc.).

Fisher’s GapAnalysis service shows the differences between a client’s current behavior and best practices in data-driven decision making. There are some patterns:

• The most common uses of business intelligence are “lookup” functions. Most companies (just over 80 percent) use business intelligence for this.
• About 40 percent of clients we’ve worked with initially used business intelligence also for estimating market size.
• The third most common use (just under 25 percent) is identifying potential customers for specific products (among suppliers) and evaluating competitors (among producers).

That said, some of the lesser-used functions of business intelligence are arguably the more valuable ones. For instance, fewer than 10 percent of companies initially use business intelligence for one of these purposes:

• Micro-segmentation: how much potential each customer and sub-segment has, and the customer’s potential ROI from buying.
• Optimization: which capital or R&D projects to invest in based on market size, value proposition, and expected ROI.
• Forecasting: how customers’ business cycles will affect demand and buying behavior.

We have observed that, while companies in the industry may access business intelligence fairly frequently, they often put it to lower-value uses. One explanation for this is that companies have trained themselves out of data-driven thinking for lack of reliable resources that support and reward such endeavors. The fact is, there are many more uses of data-driven decision making that most companies fail to exploit, yet which can significantly bring forward opportunities and drive profitability.

The following examples illustrate three important factors: the impact that business intelligence can have; how inexpensive business intelligence is compared to its impact; and that business intelligence works only when it meets certain strict requirements.

Case Study 1: Defending an M&A deal to anti-trust regulators
A large paper producer made a successful offer to acquire another company. The Department of Justice (DOJ) challenged the deal for having the potential to over-concentrate the market’s capacity. DOJ staff produced data showing a post-acquisition level of concentration in the grade that had typically caused courts to prohibit mergers.

The acquirer, however, showed with credible technical detail that many machines serving an adjacent grade had available, but hidden capacity because they could “swing” into producing the grade in question. It was argued that this expanded the definition of the grade segment (Figure 1). DOJ staff conceded the point and dropped the challenge to the merger. The cost of the business intelligence support was about five hundredths of a percent (0.05 percent) of the cost of the acquisition.

Fig. 1: Swing machines have significant hidden capacity, expanding the market definition. *Business intelligence contained in FisherSolve, a product of Fisher International, Inc.

Case Study 2: Supply chain optimization
A major consumer goods company buying hundreds of millions of dollars of corrugated boxes believed significant savings should be available through radical changes to its supply chain. An optimization model of the supply chain found an optimal selection of containerboard mills and box plants which, when implemented, saved tens of millions of dollars in production, logistics, and capital costs. The consumer goods company and its suppliers made the changes and shared the savings.

While the consumer goods company had good data on its procurement needs, the key missing piece was business intelligence detailing the production rates and costs of every containerboard machine and corrugator (Figure 2). The cost of the business intelligence support, including both the data and the optimization model, was four-tenths of a percent (0.4 percent) of one year of the savings achieved.

Fig. 2: Reliable data on third-party producers used in a supply chain optimization model produced millions in savings. *Business intelligence contained in FisherSolve, a product of Fisher International, Inc.

Case Study 3: Sales management’s resource allocation
The head of sales of a major automation supplier wanted more sales without hiring more salespeople. He believed his people could sell more if they selected more productive targets, but didn’t have a basis on which to challenge the selections his salespeople had been making.

The head of sales’ experience told him what characteristics to look for in good prospects. For example, mills with multiple DCS suppliers, older systems, operational complexity, competitive market positions, and growing grade segments were more likely to upgrade automation systems. His business intelligence resource had all this data and he used it to rank the attractiveness of every mill, directing his salespeople to the optimal collection of prospects (Figure 3). The cost of business intelligence for the project was less than 0.2 percent of one year of the new volume the company was able to pursue.

Fig. 3: Number of lines operated by each prospect that should find the supplier’s offering most attractive. *Business intelligence contained in FisherSolve, a product of Fisher International, Inc.

Case Study 4: New product development analysis
A large supplier to the paper industry developed a new material with advanced heat-transfer properties. The pulp industry represented a potential market. To decide whether or not to invest further in application development, the supplier wanted certainty that the market would be attractive, i.e., not just in terms of the total potential, but also of the value it could potentially deliver to each customer.

Business intelligence defined the size of the potential market using detail on the capacity and design of every evaporator worldwide. Interviews with a sampling of people, also pulled from the business intelligence system, defined the value of this material compared to conventional materials. Then the business intelligence system extrapolated the value to all other mills based on the system’s cost-of-production model for every mill (Figure 4). The cost of business intelligence supporting the project was less than 0.2 percent of one year of the market’s potential.

Fig. 4: Volume of potential in each type of evaporator and concentrator worldwide. *Business intelligence contained in FisherSolve, a product of Fisher International, Inc.

Case Study 5: Aftermarket sales strategy
A major paper machine manufacturer, seeing declining spare and wear parts sales, wanted to define short-term tactical actions to earn more market share and longer-term strategic moves to increase the size of the available market. The supplier defined its share of every customer’s consumption by comparing its sales to the business intelligence tool’s calculation of each mill’s consumption, which was extrapolated from details about the paper machine. Segmentation and trend analysis showed how each customer’s behavior had changed and suggested the design of a new service and pricing strategy (Figure 5). The cost of business intelligence and the related analysis was just over 0.1 percent of one year of the business’s annual sales.

Fig. 5: Market share analyses suggested sales strategies for each country. *Business intelligence contained in FisherSolve, a product of Fisher International, Inc.

In the cases described here, the critical factor was credible, high-quality business intelligence that was an integral part of a data-driven decision-making process. The leverage it brought to clients far outweighed the cost.

Companies that successfully use business intelligence to drive decision making realize greater profitability, often at the expense of their competing peers.

Purposeful data-driven decision making isn’t just a way to become a better business; it’s a competitive weapon that businesses cannot afford to bypass.

Rod Fisher is president of Fisher International, a global paper industry business intelligence consulting firm. Fisher serves producers of pulp and paper and their suppliers, investors, and end-users, and is known for helping people in those organizations make business decisions using hard data and analysis. The company is launching a powerful new generation of its business intelligence platform, FisherSolve™. Learn more at

What’s Included in Business Intelligence?
There are two types of business information: internal and external. The phrase “internal data” covers payroll, production, raw material purchases, and other activities the company can measure directly about itself. Most companies have conquered the challenge of gathering and reporting internal information with ERP, MES, and other similar systems.

“External data” describes the more chaotic outside world of markets, customers, competitors, suppliers, and regulators—a world that is complex and constantly changing. Yet it is this external world that determines most of the success or failure of the company’s activities. This kind of business intelligence can, and should be, a strategic competitive weapon.

The Problem with Unproven Business Intelligence
Let’s say you are using business intelligence to evaluate the size of a market to justify a new R&D project. Your new product would be targeted at machines possessing four characteristics that may be described by their speed, energy cost, furnish, and press configuration.

Now suppose your business intelligence resource has 80 percent of the data on each characteristic for each machine (that’s a lot; most databases have nowhere near this level of completeness). When you add up the number of machines or sum up the tons you’ve filtered out, you don’t get an “80 percent good” answer; your answer’s reliability is actually 80 percent of 80 percent of 80 percent of 80 percent— or one that is just plain wrong.

How Business Intelligence Resources Differ
In this example, you created a segment based on a number of technical details and then profiled it using a few different factors (say, breaking down the total number of machines by size, company, territory, or grade). To do that, your business intelligence tool needs to be able to perform all five of these:
• filter on every factor in the database
• sum, count, average, min, max, etc. on every database field
• produce tables, graphs, charts, maps, etc. designed any way you need
• integrate with your own information, e.g., sales history and competitive data
• roll up and drill down, show “What-ifs”, and offer complete transparency

Most business intelligence tools limit filtering to a few columns such as company, region, and grade, making it impossible to address most real-world business issues. This is because most databases were built as look-up tools; they lack the detail, completeness, and the database structure necessary to support flexible and powerful segmentation. Database design limitations also prevent systems from data integration and hierarchical analysis, which are critical to understanding your competitive environment.

Without an appropriate foundational database structure and tool design, it isn’t possible for business intelligence resources to perform at the level required for data-driven decision making.