The Reality of Real-time Performance Management

It's 11 p.m. Do you really know how well your supply and demand chain is performing at this very moment?

It's 11 p.m. Do you really know how well your supply and demand chain is performing at this very moment?

[From Supply & Demand Chain Executive, October/November 2003] The phone rings. It's the boss. She wants to know how well the company is performing right now, at this precise instant, and she ticks off a rapid-fire list of questions like she's reading them from a PowerPoint slide: How profitable is the company overall today versus yesterday? Who are the company's top 10 suppliers, and how much of the company's business are they each getting? How profitable are our 10 most profitable customers? Which plants aren't making performance targets today, and why not?

Five minutes later, without breaking a sweat, you've given your boss all the information she requested. But now an alert pops up on your computer, and you're on the phone to Topeka to see about fixing a problem with on-time deliveries to the company's top customers. A real-time executive's work is never done...

Un-real Time

In an age when the amount of data traversing the ether seems to be increasing at a rate exponentially greater than our ability to make sensible use of it all, one might be forgiven for believing that the scenario above should, at the least, be easily within our grasp, that we ought to be able to pull from our enterprise systems all the information necessary to placate an inquisitive chief and fix the problem in Topeka in something approaching real time.

But the evidence suggests that most enterprises today remain far removed from anything approaching a real-time understanding of how their businesses are performing and of the impact that ongoing events are having on their operations. For example, a recent survey by technology research firm ARC Advisory Group revealed that just 10.4 percent of companies were able to measure immediately the impact of energy cost changes on their businesses, while nearly a quarter (22.6 percent) measured that impact monthly, and more than half (53.9 percent) said they measured that impact quarterly. Similarly, the same survey showed that it takes a month or more for over 70 percent of companies to measure the impact of market price changes on their businesses.

The inability to monitor, and manage, enterprise performance in real-time can have significant consequences. Carl Lehmann, a vice president with information technology (IT) research firm META Group, notes, in a report entitled "Is the Real-time Enterprise Pragmatic?," that while networking equipment manufacturer Cisco was able to roll up its financial within a single day, helping to support the company's capitalization at the start of the tech downturn in 2000, "it failed to apply this capability to its supply chain operations, forcing it to write off $2 billion worth of inventory as the downturn worsened." Elsewhere, technology consultancy Forrester Research has reported, based on a survey of 40 financial and operational managers at U.S. and European firms, "that 63 percent of C-level executives are unable to make the right strategic decisions because of poor visibility of internal operational data." And 55 percent of those surveyed estimated that the opportunity cost of this poor visibility is greater than $55 million.

Anecdotes and statistics such as these are prompting increasing interest in real-time performance management (RTPM) within supply and demand chain circles. Why have enterprises taken notice of RTPM only of late? Steve Banker, director of supply chain solutions at ARC Advisory Group, points to two reasons. First, enterprise IT infrastructure only recently has evolved to the point where it is becoming technically possible to pull together data from disparate corporate systems at or near a real-time pace, something that the batch-driven enterprise resource planning [ERP] and other legacy systems could not have supported. "But the other piece," Banker continues, "is that the world has gotten a lot more dynamic. Things seem to be happening in the external environment at a faster pace, and the static planning we used to do in the past is really less and less tenable."

In addition, from a supply and demand chain perspective, the push to "get real-time" can be attributed in part to such trends as the drive to reduce costs by moving to just-in-time operational models, according to Mike Schiff, vice president of data warehousing and business intelligence at IT consultancy Current Analysis. "In the supply chain world, you need to get the right parts or the right product to the right place at the right time," Schiff says. "In order to do that, you need information that's timely. Just-in-time inventory requires just-in-time information  it's basically that simple."

The "Real-time" Technology Stack

For those enterprises interested in becoming more real-time-oriented, the bad news, analysts say, is that no single, "silver bullet" RTPM solution exists to get you to that goal. "We don't see that you can go out and buy this solution from one supplier," Banker says. "What we really see is a series of components that need to be intelligently integrated together."

The good news, analysts agree, is that major enterprises likely have many of those components already in place. For instance, in ticking off the components that he believes are necessary to enable RTPM, Banker starts at the bottom of the technology stack with the various operational data sources that are already are humming away at most companies, including an ERP system, manufacturing execution and automation systems, inventory management systems and customer-oriented systems. These may also include such new technologies as radio-frequency identification (RFID) systems or global positions systems (GPS) that allow companies to track inventory in real-time across the supply and demand chain.

Next, ARC believes that companies will need to implement, if they have not already, new, very granular costing and revenue models that can help executives better understand the impact of events within, and external to, the enterprise, as well as the consequences of their decisions. "For example," Banker offers, "if we're trying to understand why Customer A is more profitable than Customer B, we really have to understand the activity-based cost around the particular value-added services we provide to Customer A versus Customer B. We cannot use standard costing, which is what most ERP systems are based upon."

On top of these models, enterprises will need a host of decision-support tools, Banker says, including solutions for financial planning, simulation and optimization, real-time costing, operational support and dynamic target-setting. And on top of it all, ARC sees a role-based dashboard available through a Web portal so that all stakeholders can look at the information relevant to their functional role in a way that's meaningful for them. Note that these would not be static dashboards presenting purely historical data, but dynamic workspaces that not only offer up-to-the-minute information but also allow an executive to drill down into the data behind higher-level metrics to understand the root causes of particular issues.

Other analysts offer similar laundry lists of technologies that can be cobbled together to create a viable solution, adding such nuts and bolts as extract-transform-load (ETL) tools, a data warehouse infrastructure, financial consolidation applications and business process management tools.

"So there are a lot of components," Banker concludes, "and I don't think that there is any one company that has rolled those components together in a way that's designed specifically to hit the vision of real-time performance management. But that said, with a service-based architecture and the enterprise application integration (EAI) tools that we have, you can get there with the tools that exist in the marketplace today."

Setting Real-time Targets

Among the technical challenges in achieving real-time performance management, data quality stands out as perhaps the toughest nut to crack. "People are going to have clean up their data," Banker says flatly. "They're going to have to come to common metrics that are defined the same across the corporation, and maybe even across an extended supply chain." Because enterprises may need to patch together their data quality tool set using a variety of solutions to address different components of their RTPM infrastructure, the price tag could be hefty, but the alternative, as Banker puts it, is "making bad decisions more quickly than you used to."

Naturally, just having all the right technology in place won't get an enterprise to real-time performance management. For instance, companies will need to decide what metrics they will monitor in real time. The danger is as much in measuring too much as it is in measuring too little, lest a company become overwhelmed with data. "Only some of an enterprise's business processes will require real-time awareness," writes META's Lehmann. In particular, the analyst points to those processes "that require constant status of inventory availability, order management, logistics, just-in-time manufacturing or build-to-order/customized products that must tightly adhere to time-to-market or delivery commitments."

More broadly, Banker suggests that RTPM may require that enterprises rethink the way they set their performance goals, moving from static targets (set at the beginning of the year and frequently obsolete the minute they're finalized) to dynamic targets that can change to reflect the evolving economic and market realities of the day. The twist here is that while enterprises with static targets typically cascade performance goals down the organizational chain, with real-time performance management it actually could be easier  and more effective  to build targets from the bottom up. As an example, Banker cites a large discount retail chain that started its goal-setting process by determining the business drivers that made each store profitable; setting targets, metrics and incentives to promote the appropriate behaviors at the store level; and then aggregating those targets (and assigning the appropriate metrics to be monitored) to derive goals for successively higher levels back up the organization.

Working with dynamic targets may be more difficult for manufacturers than for retailers, but it is possible. Explains Banker: "In manufacturing, dynamic targets are often operationalized around flexible budgeting. The idea is to set cost targets  for example, cost per pound  but then to break the budget for a plant into fixed versus variable costs. Everyone realizes that the fixed costs, like rent, are not going to change, but variable costs, like energy, do. Managers automatically have cost targets raised if energy costs go up, and their cost targets are automatically lowered if the cost of energy goes down."

As the example of the retailer above implies, the move to dynamic targets within a real-time performance management framework may also require companies to rethink how they set employee incentives, for instance, moving to a system of compensating sales staff based on profits rather than on revenues, and then ensuring that the RTPM dashboard reflects the proper metrics so that employees understand how well they are meeting their personal targets. Shifting incentives could well be the most difficult "cultural" challenge in making a real-time performance management initiative successful. However, as Monica Barron, AMR's research director for human capital management, said in discussing dashboards at the consultancy's spring executive conference earlier this year, "People are coin-operated. [They] will focus on what you are incenting them to do."

The Real-time Reality

In addition to the above challenges, Banker suggests that the disconnects between functions within the enterprise represent a key barrier in the way of real-time performance management. "Logistics and distribution managers have begun to operationalize cross-functional cooperation in the supply chain, and they've also begun to do this a little bit with sales and marketing around processes like collaborative forecasting," the analyst says. "But there is still a huge disconnect between sales and marketing on the one hand, and the supply chain folks on the other, because of the sales quotas set in the budgeting process."

As a result, salespeople may ship out product at the end of the quarter just to meet quotas, knowing full well that the products will be coming back from the customer. "We need to find a way to better integrate the functions so that we're doing a better job from a company-wide profitability perspective," Banker asserts. RTPM could serve as a catalyst for this type of integration by providing sales staff with real-time insights into the cost of servicing different clients, allowing them to understand how each order would affect the profitability of a given customer. And again, the company would have to provide incentives to sales staff based on profitability rather than on revenues.

To date, Banker says he has seen few companies moving determinedly toward real-time performance management. And given the technical and cultural barriers to achieving RTPM, no company may ever achieve "perfectly real-time" operations. But as companies contend with the growing interconnectedness of world economies and continued political and economic volatility around the globe, they may find themselves increasingly compelled to adopt aspects of the real-time enterprise out of competitive necessity.

As Stanley E. Fawcett, of Brigham Young University's Marriott School of Management, and Gregory M. Magnan, of Seattle University's Albers School of Management, wrote in a paper entitled "Achieving World-Class Supply Chain Alignment: Benefits, Barriers, and Bridges," "One common tenet [of supply chain literature] is that competitive success depends on managers' ability to recognize changes in the competitive environment and then to structure organizational and, where appropriate, supply chain resources to effectively meets customers' real needs." RTPM promises to empower managers and executives with the tools they need to recognize those changes  and the information they need to ensure appropriate and timely responses.