At various times, at conferences and in conversations, one often hears the usual cynical statements regarding forecasting practices, software and technology: "Yes, so and so company is in trouble," or, "Forecasting yields small returns for the efforts."
Forecasting, in many businesses, has been so problematic that many companies have significantly changed their business models, products and so on. In essence, as Stuart Smith, past vice president of supply chain with Dell Computer, held, if you can't forecast the business, make the business more forecastable. But, the market ultimately decides what you sell, and the gains for even small improvements in forecasting yield big returns in corporate performance.
Part of the problem, though, is that firms look at forecasting as an isolated activity rather than part of an overall set of demand management practices. And as the boundaries of business and technology expand, so do the process innovations, creating a rich set of demand management practices across the whole supply chain, the whole life of the product and the time zones of integrated business events that drive to the outcomes we seek: customer loyalty and profitability. Sounds good; where do I start?
Not that you have to be reminded, but business has significantly changed in the last few years. The four-wall enterprise has yielded to a more dynamic and global model. This means that the solutions designed to support the old model probably won't work so well in the new world (license software was built on the monolithic definition of the enterprise, but that discussion is for another time). At the same time, we are seeing several key technologies come to fruition that will change the way most enterprises do computing, and the new models can significantly revitalize our demand management practices.
In this article we will discuss these changes in business and technology and describe the significant expansion of demand management practices that work. In addition, we will discuss practices that your firm should be developing to significantly improve business performance, based on ChainLink's Future Forward models, which have been validated through our research.
The Digital Essence of All Things
As we globalize, enterprises are moving beyond the virtual models of the late 1990s and are managing their trading partners differently, both on the supplier side and the channel side. These "federated" models rely on a high level of integrated information technology to thrive. Bill Gates' statement regarding the debate on DVD formats — "Understand that this is the last physical format there will ever be..." — makes us think about how we leverage the digital universe to better manage our business. Information availability through network applications, radio frequency identification (RFID) and so on is expanding visibility and at the same time reducing the cost of information technology. (See Figure 1.)
Figure 1 
The creation of new dynamic business structures (federated models) that are enabled by, and rely on, these ubiquitous platforms points to a growing capability to understand and access market intelligence for all businesses. In the past there was very little access (without huge investments) to intelligence to understand markets and consumer behaviors. This led to an over-reliance on channel partners, who frequently had no motivation to provide these data. Today this is a huge inertia issue in most industries; face it, when was the last time your firm did its own sensing about what was actually happening in the end-customer market? And if your company has conducted some efforts in this direction, is it a process?
The reality is that we are master of very little of what it takes to excel at demand management. And the diffusion of the enterprise definition, which leaves many firms as mere influencers, can exacerbate the problem. In the past, people may have doubted your numbers, but they were beholden to follow your dictates. Today this might not be the case. Therefore, reaching out to markets, sensing and visualizing them all should be front and center in your company's demand management practices. Creating these models allows you to look at several key issues, such as in which markets your company can make money, what products go into which channel partners, or how fast product lifecycles need to turn to effectively compete.
The next digital zone is item-level identification through RFID and related systems. Though this is new stuff, many firms can get in on this action already. And linked to this information will be a cornucopia of opportunity, such as in-store operation optimization, more effective promotion management and better management of returns. (Many firms do not effectively use point-of-sale data today, so we recommend enhancing your utilization of these data now.)
Make It a Process
Most forecasting exercises are studies in introspective "spreadsheetsthenics" (spread sheet prowess) vs. infusing the system with fresh intelligence from the outside (markets, customers, trading partners, etc.). Looking at most information technology budgets will reveal a huge pile of cost locked up in legacy IT systems (with or without enterprise resource planning systems — legacy support statistics by various analysts put the number at 50 to 80 percent of budget) and not much investment in technologies that can help grow the business. Subscribing to fresh data sources, market analytics, visualization (data and tools) and understanding trends usually consume less than 1 percent of IT budgets, yet they are directly in line with growth opportunities.
Now from markets, let's look at products. Product demand characteristics change over the lifecycle of the product. Trends and fashion changes turn on and off the spigot of demand. Auto and high-tech pile on features to extend the life of product. Product launches, promotions and other stimulants to the market introduce variability. You need to account for these and overlay traditional patterns like seasonality with lifecycle patterns. Over time, your knowledge of how to manage these will improve.
Managing across the chain and creating responsive and dynamic processes with understandable metrics is critical. Looking at metrics and simple things like forecast vs. actuals will allow you to learn and improve. Evaluate and share these metrics and learnings across the time zone of business events (and your trading partners). What assumptions went into the early phases of market discovery through end of life? What can be learned from these? Very few of these are practiced in a systematic way.
Expanding the Practices of Demand Management Across the Chain
All businesses have to compete for scarce resources — customer's wallets, scarce resources from supply markets — and for time. As we import more, critical time is lost in the ocean lanes. Understanding longer-term needs as well as methods to become more responsive (a worthy subject for another time) becomes even more critical, not less. So giving up due to cynicism is not the answer. What is? Using a broader set of methods and tools appropriate to your markets.
As we rely more on our trading partners, changing the nature of relationships becomes critical. Today's approach becomes inclusion, building a collaborative framework, creating a consensus and increasing sharing in the chain. Adversarial relationships build shortages in the chain – and more costs. As companies become more sophisticated in their process, information consumption and analytics, they understand the impacts and drivers to cost. Ultimately, a lack of collaboration will cost you more in terms of time to market, product costs, expediting costs or shortages at the shelf. Some manufactures of late have fired their retailers since they understand now that they just don't make any profit from certain customers, and no amount of scale can reduce the effect. Better to be a bit smaller and very profitable than large and go out of business. Adding to the arsenal is the ability to understand and manage the risks involved. Risk management becomes a key element of today's intelligent enterprises.
A full treatise on demand management practices could fill books, but included here (see Table 1) is a framework of practices or elements for your consideration.
Table 1
Conclusions
Remember there are many sources of uncertainty in supply chains, and these sources vary over time — maturity of, or changes in, relationships with the customer; market knowledge; product lifecycles; as well as the impact of competitors on demand – so demand management is a dynamic set of practices.
Some key thoughts, though, with which to move forward:
- Forecasting is the middle, not the beginning, of demand management practices.
- Planning assumptions have to be codified in order to learn and improve the process.
- Performance objectives and assurances need to be shared, monitored and refreshed in the n-tier chain.
- Sensing techniques can yield an improved competitive position. Get a jump on what the market will want next, or now, and move that intelligence into your product design and supply chain.
- Risk management techniques are now beyond "scientist methodology" and can be taught and implemented.
And most of all, in the new dynamic global business models, getting your trading partner into the mix is critical to success, whether customer or supplier. Two heads are smarter than one.
