As a commodity business, United States Sugar Corporation is used to operating at the whim of Mother Nature. While they might not be able to fool her, new tools and techniques have helped executives at U.S. Sugar's Clewiston, Fla., headquarters -- as well as farmers in the fields -- plan and respond to the vagaries of a highly unpredictable operating environment.

As the nation's largest sugar producer (with an estimated annual revenue of approximately $500 million), U.S. Sugar has grown into a complex conglomerate of agribusiness companies that manages more than 194,000 acres in Florida, including the annual production of 800,000 tons of raw sugar and 120 million gallons of orange juice (from 15 million orange trees), as well as sugar by-products (such as molasses for animal feed) and other commodities.

The farming business is notoriously slow at adopting new technologies. Even though U.S. Sugar provides 10 percent of the nation's sugar, our infrastructure frankly has not kept up with the times. One of our two sugar mills dates back before 1928, and the "new" mill was built in 1959. And while it's true that the technology of grinding cane hasn't changed much over the last century, the dissemination of production data certainly has.

In the late 1990s, in anticipation of Y2K, the U.S. Sugar management team decided to seek out new technologies to modernize the gathering and dissemination of production information. Bringing U.S. Sugar into the 21st century wasn't going to be easy. Technologically speaking, we were contending with a slew of disconnected legacy systems built 30 years before, and most of the production data coming out of the mills was collected manually. The rest of it resided in the collective heads of generations of farmers so attuned to the rhythms of the earth that decisions on when to plant or harvest were as natural as, well, the weather. As a commodity business with little control over most things other businesses take for granted (for example, the government sets the price and the amount of sugar we can sell), we felt it imperative to find a better way to control costs. Historically, the nature of our business has caused us to react to -- rather than anticipate -- events that we can't control, and we needed to find a way to rectify that.

BPM Beginnings

In the summer of 2000, a new CEO and CFO -- neither of whom had extensive experience in the sugar industry -- joined U.S. Sugar at a time when the industry was being challenged by historically low sugar prices due to a surplus of subsidized foreign imports. As a result, sugar-processing plants across the country were closing. New management could afford to rely only on hard-and-fast data and analysis, not on anecdotes. This was a major culture change for the company, and it resulted in an inordinately high volume of financial information requests.

Although we, as a company, like to talk about "working smarter, not harder," the accounting staff ended up working harder, not smarter. Not only did we experience volume overload, we were constantly plagued with clerical errors from rekeying data, conflicting data from disparate databases, and inconsistent definitions of the same term across departmental lines. With at least three different ways to measure a ton of raw sugar, we would have different results for the same metric depending on who performed the calculation. As with a lot of ERP systems, we had a lot of transactional data but no good tool with which to turn the data into information. We spent a huge amount of time and resources on data gathering and reconciliation, which left little time for the actual analysis.

U.S. Sugar happened on the doorstep of business performance management (BPM) by accident. Several finance and IT staff members had been attending various industry and professional conferences, and all came back talking about software solutions that could marry data from different databases to better enable us to plan, report, analyze, and manage performance. Struggling to resolve various versions of the corporate truth and meet management's pressing information and analytical needs, my team and I strongly advocated investment in a BPM solution (although the software didn't yet go by the name "BPM").

Our primary goal was to provide both the executive and operating management teams with information they needed to run the business. The information had to be timely, accurate, and clearly defined. It needed to be at their fingertips instead of down in the bowels of accounting. A new solution had to be able to report our results from a single database where everyone could easily access the same information to draw accurate pictures of business performance. Such a solution -- if it could be implemented affordably -- had everyone's support.

The Evaluation Process Begins

One of our first priorities was to achieve departmental buy-in across the enterprise before the software selection process began. Taking a cross-functional approach from the project's infancy, finance invited all departments -- IT, agriculture, facilities, equipment maintenance, purchasing, sugar processing, citrus operations, and refinery -- to respond to two basic questions: "What are your needs?" and "What kind of relief are you looking for?" Based on group input, we came up with selection criteria that centered on application functionality, technical platform, data security and integrity, flexibility and user-friendliness, price, and the financial viability of the BPM vendor. Secondary to specifics was the need to internally drive home the fact that the chosen solution would not be exclusively an accounting tool. In fact, the chosen solution had to clearly demonstrate that we didn't need dollar signs in front of the data in order to provide value to the organization.

Ultimately, we needed a solution that would bridge disparate databases, required little or no IT intervention for developing reports, fit our current IT infrastructure, and was affordable.

We did not craft a formal RFP or vendor checklist for the evaluation process. Rather, a team of IT and accounting staff researched software vendors on the Internet and at industry events over a six-month period. A short list of three vendors for further evaluation was established. Each vendor was asked to provide references for both telephone interviews and site visits. More in-depth evaluation was conducted when each vendor demonstrated its product before our larger cross-functional evaluation team by developing reports to our specifications, using our data. When each vendor completed its scripted demonstrations, the U.S. Sugar team requested ad hoc modifications to judge flexibility and ease of use.

Satisfied that we understood each solution's functionality after these meetings, our team narrowed the selection criteria to ease of use and price. Our choice: OutlookSoft's Enterprise Analytic Portal. EAP best met our unique needs: It had one module and no add-ons, it was Microsoft-centric, it fit well with our IT infrastructure, it did not require IT intervention to generate reports, and it cost only 25 percent of the total price of the next best solution when implementation and training were included.

Harvesting Early BPM Results

With finance taking the lead, U.S. Sugar began its implementation in January 2002, integrating the system with our legacy mill production and agriculture databases, our ERP systems (MLS and OPEN4 systems, all running on Progress databases), and a SQL-based fixed-asset management system (see exhibit 1).

The implementation began as a pilot program. Even though the evaluation and selection process was cross-functional, for cultural reasons we concentrated on financial reporting first. Empirical evidence demonstrating how efficiencies had been gained in the financial statement process would strengthen the "take-up" of the system in other areas of the organization.

The team spent four months mapping the data from our ERP system and reconciling the results. We loaded the current year (FY2002) and the prior year at the account level, so no transactional data was loaded. The April financials were published using the tool in May, and the time it took to prepare them had been reduced from 10 days to two days.

We then tackled the budgeting application. Our budget cycle normally started in April, with the spring crop estimate, and ended in a flurry of activity in late September, just in time for the October year-end. We develop and consolidate the operating budgets of 46 entities and 288 cost centers, with 171 expense categories, plus the results of our citrus and railroad subsidiaries. Previously, all of this was done in Excel. The process easily fell victim to "version confusion," broken links, debits not equaling credits, and other assorted nightmares that had to be resolved in accounting.

The budgeting worksheet formats were migrated over to OutlookSoft EAP, where templates were published for use by operations. The roll-ups were developed in the system and became almost entirely automatic. Changes in global expenses, such as employee benefits, were populated into everyone's budgets with a couple of keystrokes. Monthly financial statements, including the P&L statements, balance sheet, and cash flow statement, were developed and tied together.

The benefit of our move to BPM was demonstrated by reduced time dedicated to these essential tasks. During the first year, we took three weeks off of the budgeting process. We reduced it another seven weeks during the second year. Further reductions are expected this year. Obviously, the shorter our budgeting cycle becomes, the clearer the picture we have about crop size, which correlates significantly with our financial results.

Other applications under development in the financial area are a capital budgeting and reporting system and cash forecasting. Dashboards for various executives are also on the plate.

The finance managers in operations worked with us on budgeting and variance reporting; the biannual forecasting process went from a finance function to a participatory process involving every department, and the time spent on it was reduced from one month to one week. Colleagues across the enterprise began to see value in the implementation, as there was a lot more transparency and ownership of the numbers. Department heads could understand the key factors in their budgets and forecasts because they themselves helped create them. Knowing that they could access the information they needed to run their business better without having to go through an intermediary fostered solution acceptance throughout the organization.

Extending BPM to Operations

In order for our BPM solution to succeed enterprisewide, we had to address what came to be one of our biggest pain points -- definitions of common terms. There were as many definitions of a term as there were departments using it. We knew it wouldn't be enough to just capture, consolidate, and distribute information from various legacy systems; it was vitally important that everyone in the organization understand the definition of terms used to view and analyze that data. We then identified and outlined the points where data differences could arise and crafted a standardized dictionary of definitions to be used in the development of all applications.

In addition to creating a dictionary of terms, we devoted substantial time to determining our expectations -- what we really wanted the new BPM solution to accomplish -- by carefully evaluating business processes across the organization. A thoughtful, measured analy-sis of how we planned to use the new solution was fundamental to the success of the implementation.

After seeing marked results within the financial group pilot, the agriculture department developed an application to automate the entire maturity curve process for U.S. Sugar's annual $100 million sugarcane crop. Traditionally, the data regarding countless varieties of sugarcane and soil types was gathered manually and keyed into spreadsheets for each of the company's nine farms in order to determine optimal sugarcane harvest and planting times. The information was vital but it yielded a static snapshot that aged quickly as environmental conditions changed. After developing the dynamic models within the BPM system, a farm manager (who is just an average user) can now select his farm, look at his maturity curves -- in total, by certain varieties, and by soil types -- and determine his harvest and subsequent planting schedules.

This system has yielded both savings in terms of time spent on analysis and productivity gains based on better decision-making. In fact, the success of the application is measured by the fact that more of our farm managers and coordinators are requesting it. That's a huge step for a culture traditionally very distrustful of computers and software. Farmers spend every day close to the land; they're not known for relying on modeling. But with this new tool, they can now maximize the productivity of their land and create optimal conditions to increase crop yields and manage fallow acreage.

The How-To on Implementation

Working with our BPM solution provider's consultants to guide us in designing the applications and reports was a critical factor to our success. It was required that the business processes and information needs of all system users be mapped out in detail ahead of time. That allowed us to configure a comprehensive plan, ensuring that we included the right information from the start and that we had the right people on the team.

Knowledge transfer was initiated as soon as the process got under way, so our reliance on the consultants was minimized for our current applications, allowing them to focus on new applications in other areas.

User acceptance testing (UAT), inspired by our consultants, was an integral part of our implementation. At various times during the system's development, end users of the application, at various levels, were shown the product to date. This ensured we were able to maximize functionality and ease of use, while also improving user buy-in, which is critical to the success of any project.

We have begun taking our BPM solution to the next step, working to develop predictive analytic models that provide even more insight and accuracy into our forecasting and analysis of the business, promoting keener understanding of the complex cost structures surrounding land usage, equipment, time, labor, and materials. These predictive capabilities promise to reduce the amount of time spent gathering and analyzing data by creating an auto-forecasting system applicable to everything from crop yields and weather to operational requirements. Twenty years of historical agricultural data on fields, varieties, soil types, fertility, weather, and yields will be loaded in order to determine the most profitable combinations and identify potential problems. Our vision is to unlock previously hard-to-quantify information -- a farmer's inherent knowledge of his fields, the weather, and his crops -- and capture it in a form that will influence, educate, and aid current and future stewards of the land.

The same concepts are being applied to the mill side of the business, where manual processes in place since the 1930s are in the midst of a long-overdue update. Our mills will be able to look at sucrose levels and other quality characteristics from the sugarcane grind, as well as consumable materials, labor, and equipment, to help establish where operations are running smoothly, where bottlenecks are, and where there is potential for cost savings. These financial and operational processes are also being used to more effectively manage our environmental concerns -- activities such as fertilization, irrigation, and cultivation. Considering the proximity of U.S. Sugar's land assets to the Florida Everglades, we have already pioneered integrated pest management to minimize the environmental impact of pesticides. Using these new data collection and analysis tools allows us to optimize fertilization and cultivation rates in an effort to further minimize their environmental impact.

The Final Analysis

U.S. Sugar's BPM initiatives remain a work-in-progress, but the organization has clearly benefited from the changes this initiative has already brought. It goes without saying that we learned the value of business processes that support data collection. That being said, organizations should not hesitate to reengineer their business processes to achieve better results. An effective BPM solution will not only utilize all the information available within the organization in order to make better, more timely decisions, but also will identify where business processes need to be improved or changed. As our results at U.S. Sugar prove, an organization's willingness to change, to learn from past experiences, and to clearly set expectations for improved business management processes will dictate the level of results achieved.

About United States Sugar Corporation

  • U.S. Sugar is one of America's largest diversified, privately held agribusiness firms. While the company is best known for its sugar and citrus, it also owns a short-line railroad, South Central Florida Express.
  • U.S. Sugar owns 194,000 acres of some of the most productive farmland in the United States in a fertile crescent from the southern shore of Florida's Lake Okeechobee and extending into Glades, Hendry, and Palm Beach counties in South Central Florida.
  • U.S. Sugar's primary business is growing sugarcane and producing sugar -- some 800,000 tons each year. The addition of a sugar refinery in 1998 allows the company to refine the raw sugar that it produces -- sugar that previously was shipped to out-of-state refineries for processing.
  • Citrus is the company's other main business. At Southern Gardens Citrus, 120 million gallons of orange juice are produced annually. U.S. Sugar is the largest supplier of private-label, not-from-concentrate orange juice in the United States.
  • U.S. Sugar is partly owned by its employees through an ESOP and members of the Mott family and related foundations and trusts. Its estimated annual revenue currently is approximately $500 million, and it employs 2,500 people.

BPM Implementation/Administration Timeline

  • January 2002 -- Software is loaded.
  • April 2002 -- Budget application development and rollout begin simultaneously.
  • May 2002 -- April financial statements are prepared and published using OutlookSoft; time is cut from 10 days to five days, later to be reduced to two days.
  • September 2002 -- The 2003 budget package is finalized for presentation to the board.
  • June 2003 -- Capital budget and reporting application are started.
  • June 2003 -- The 2004 budget process is started.
  • September 2003 -- The 2004 budget is available for board review (a reduction of seven weeks).
  • December 2003 -- User acceptance testing for capital application is held.
  • December 2003 -- Devel- opment of agriculture services labor application begins.
  • February 2004 -- Expected rollout of capital application.
  • February 2004 -- Expected rollout of agriculture services labor application.
  • August 2004 -- Expected start of 2005 budget process.
  • September 2004 -- Expected completion of 2005 budget for board review.

Tangible Business Benefits from U.S. Sugar's BPM System Implementation

Approach Result
COST

$660,000

$360,000 in implementation and consulting

$300,000 in software

Minimal cost impact
YEARLY BUDGETING CYCLE Budgeting cycle reduced nearly 50 percent (from 20 weeks to 11 weeks) Better business reaction time; finance now focused on analysis and impact (not process mechanics)
FORECASTING CYCLE (2 TIMES/YEAR) Forecasting time reduced 75 percent Reduced direct cost; more accurate forecasts; high degree of employee ownership
REPORT AND ANALYSIS

Self-generated reports and analysis (unlimited)

User-defined reports in addition to systemwide reports

Clearly defined and agreed-to terms and definitions

Employee empowerment; higher degree of ownership

More informed user decisions; faster decision time; shorter market reaction time

Consistency in interdepartmental reports; higher confidence in decisions; faster decision-making

IMPLEMENTATION PROCESS

Democratic approach (all departments)

Participatory design and implementation

In-depth evaluation of user needs and processes

Enterprisewide buy-in and support

Better skills for ownership; self-management; continued improvement

High user acceptance across the organization; solution capabilities aligned with department needs; more productive work environment; employee sense of empowerment

Ellen Simms has served as corporate controller at U.S. Sugar since 1996 and has been with the company for more than 20 years.