Intelligent Enterprise

What's your plan? Your company's success could hinge on how different departments and business functions answer that question, and how well executives synthesize the answers into an overarching business strategy. Unfortunately, few things are more painful than the planning ritual. And with this pain, there's usually no gain — only confusion and market opportunities lost because your organization can't react quickly to change.

Performance management initiatives shouldn't stop with the establishment of metrics that assess what's happened in the past. Most important is to use the information to guide your organization toward a more intelligent plan, one that supports key departments and business functions as they try to align efforts.

Getting to this vision requires organizational and technological change. Today, departments typically use spreadsheets to create plans representing their own set of business assumptions and market perspectives. Sales develops a revenue forecast by region based on the current pipeline and next year's objectives. Finance creates an operating plan based on current-year budgets and next-year projections. Marketing makes forecasts for new and existing product lines based on research. Manufacturing constructs a demand forecast to drive a supply outlook based on history and future business expectations.

How do you bring these plans together? Each one offers a different perspective:

• Sales forecasts are revenue numbers in dollars by regions and customers
• Marketing forecasts are in dollars and unit sales by product lines
• Finance forecasts are in dollars by internal organization
• Demand forecasts are in units by SKU numbers and their components.

The practice at most companies is to let finance organizations reconcile the spreadsheets manually, identifying gaps. This process involves days of back-and-forth e-mails, meetings and phone calls, which should give finance departments a better understanding of underlying assumptions. Departments are then asked to tweak their plans to make sure they fit together within an acceptable margin of error.

Manual reconciliation is not only error-prone, but it's also absurdly slow. As a result, organizations update their plans infrequently — once a quarter at best. The business environments in industries such as high-tech manufacturing change much faster than that. Forecasts made last quarter may no longer be relevant — and in fact could blind companies seeking clearer visibility into business performance. Organizations are forced to use numbers that could lead to lower-than-expected revenues and profit margins, missed opportunities, lost chances to write off inventory and inaccurate customer service metrics.

Needed: Enterprise Vision

Businesses need a new approach that simplifies planning and enables them to set a more frequent, continuous agenda. This process should enable companies to create operational plans quickly and then leverage technology to refine them easily. The technology solution should incorporate each department's plan as well as the underlying assumptions about projected sales, inventory or other critical elements. The result should be a single, shared view of the plan ahead.

An enterprise approach shouldn't prevent each department from seeing the shared plan from its own perspective. Manufacturing should still see detailed component demand; sales should be able to view expected revenue by region; finance must easily calculate revenue and margins by geographical and operating entities. However, the enterprise planning process should refine the blueprint on an ongoing basis — even once a month — to bring forecasts in line with changing business conditions.

The key step in creating a single, shared agenda is to establish an analytic architecture that lets the same plan be viewed along multiple dimensions at varying detail levels. In other words, organizations must create an enterprise plan of record: a single repository of past, present and future forecasts, plans and assumptions from across the enterprise. This record would reconcile disparate aspects of each functional document into a single model, while retaining the integrity and detail needed by departments.

With an enterprise plan of record, each team could view the road map on its own terms — such as units, dollars or margins — from lowest product detail to top-line revenue forecast. Finance teams could continuously evaluate alignment with business plans for revenue and forecast outcomes, create alternative scenarios and conduct modeling exercises. At the same time, assumptions about pricing strategies, product mixes and sales forecasts could inform business blueprints focused on maximizing profit, not just revenue. With a consolidated operating plan, executives and managers would more consistently drive results, measure performance and ensure compliance with regulations.

Address Key Requirements

In addition to implementing an enterprise plan of record, a shared approach must also support the "arithmetic" capabilities that rigorous planning demands. These functions include the ability to use adaptive analytics to select the best combination of forecasting algorithms and to incorporate causal and qualitative information. As data flows in, adaptive analytics update the model. In manufacturing, for example, an adaptive analytics engine is valuable because demand forecasting must address configured products, multitiered sales channels and other changing factors. These changes must be reflected in the analytic model as soon as possible, if not in real time.

A shared approach must interpret information coming from sales pipelines, which are good indicators of future demand. Data from the pipelines usually exists in an aggregate form, and to be useful for planning it needs to be translated into demand units. Sales pipeline data also contains the natural biases of individual sales representatives, which contributes to inconsistency across the pipelines. A planning engine must incorporate heuristics and intelligent filtering capabilities to address biases and inconsistencies.

Product introductions present another manufacturing-specific challenge. To assimilate new products into the shared plan, the engine must support the variety of product life-cycle and marketing models used in manufacturing. These models typically represent assumptions about product release dates, order projections, promotions, pricing and other key issues.