The Intelligent Food Chain

Welcome to The Intelligent Food Chain - a series of news and opinion pieces where we explore how business intelligence (BI) can be used in the food value chain. Please check back often since we post a new article every month.

This summer The Wall Street Journal ran an article relating that many retailers are cutting back on their assortments, the number of items they carry in each category.  Walgreen cut the number of superglues it carries from 25 to 11.  Wal-Mart cut the number of tape measures from 24 to 4.  Kroger wants to reduce its number of breakfast cereals.  And so on.

The article quoted representatives from several food/CPG manufacturers and several retailers.  There was significant diversity of opinion about whether more items or less items is better for the consumer, which channel participant (manufacturer or retailer) benefits, which companies among the manufacturers and the retailers benefit from smaller or larger assortments, whether the market-leading brand in a category prefers a larger or smaller assortment, and so on.

The interesting aspect of this debate is that the participants in it all base their arguments on analysis of the same data (retail point-of-sale data), the availability of which has become fairly ubiquitous.  “Single version of the truth” has long been a holy grail for business intelligence (BI).  But what this story indicates is that looking at the same data doesn’t necessarily mean everyone will arrive at the same conclusion.  It seems that cognitive filters (defending past decisions, what we want the answer to be, and other perceptual biases) can overwhelm the facts, and enable the same data to tell many different stories.

As designers and builders of BI applications, can we do anything to prevent or overcome situations like these?  The short answer is, No.  Even if we are painstaking in constructing a common data repository containing a single version of the facts, “single version of the facts” is not the same as “single version of the truth,” because truth is, at least partially, a function of the interpreter of it.  Even the context with which we surround facts in the presentation layer reflects what the people from whom we elicited the requirements considered relevant and useful.

We can, and must, provide our constituencies with accurate data – correct facts, if you will.  But how they interpret those facts and what decisions they make with them is something that they must own; we cannot.

A recent article in The Wall Street Journal (August 10, 2009) tells how supermarkets in the U. K. are flourishing, despite the recession there.  The top four chains in market share – Tesco, Asda, Sainsbury’s, and Morrisons – all show growth in sales and operating profit.  How are they accomplishing these good results?  By reacting quickly to the economic environment.  The chains have reduced prices, launched new budget brands, offered cheaper cuts of meat, and advertised aggressively.  Some have even changed their point of differentiation. Sainsbury, for example, is known for its quality food but has changed the emphasis of its advertising to low prices.

How does business intelligence (BI) support these new competitive efforts and this success?  By providing frequent, timely information about product pricing relative to the competition.  The article notes that Asda collects lists of prices from all grocers’ Web sites and provides them to executives every weekday morning.  Pricing analysts review these price comparisons and send price adjustments to individual stores.

A competitive strategy based on quick reaction to changes in the environment and actions by competitors requires sophisticated, continual, and low latency monitoring of that environment and those competitors.  A purpose-built BI application can enable such monitoring.

The most popular article in The McKinsey Quarterly in the second quarter of 2009 describes how the recent economic environment is forcing companies to change the way they do strategic planning.  Three tips given are: be realistic about scenario planning, intensify monitoring, and look beyond the crisis.  Two recommendations from the article are:

“Since the effectiveness of such a strategy depends on an organization’s ability to adjust rapidly . . . managers must identify and intensively monitor key indicators suggesting which scenario might unfold.” (italics added)

“This year’s planning process should also generate unusually specific plans to monitor the performance of suppliers, customers, and competitors . . . the most entrenched incumbents can plunge into financial distress with dizzying speed.” (italics added)

The gist of the article is that both the rapidity and magnitude of change in this recession can be much greater than most companies’ planning processes are designed to handle.  Companies must monitor the environment for such changes much more intensely than they have before, else they might react too slowly, or not even realize that they need to react until it’s too late.

“That makes sense,” you might be thinking, “but our strategic planning process already includes a review of environmental factors.”  I’m sure that it does, but your strategic planning probably is an annual process that takes several months to execute.  The environmental factor review probably is done at the beginning, so that information is two to three months old by the time executive management agrees the final strategic plan.  Let me reset your understanding by asking, what if you need to monitor the environmental factors relevant to your business every week or every day?  How would you do that?

The more frequently you must measure and the lower latency (the time between when an event occurs, and when you know about it) you need, the more you need an automated solution.  Manual collection and analysis methods simply won’t support the process tempo you require.  The automated monitoring solution for you likely will be a business intelligence (BI) application.

To apply this concept specifically to food manufacturing, we first need to consider the external factors of interest.  Significant items might include:

• The price per barrel of crude oil
• The market price of agricultural commodities that are ingredients in its products
• Retail off-take at the supermarket shelf
• The frequency with which people eat out and the average check dollar value
• The unemployment rate
• Large share price drops for customers, suppliers, or competitors

In most cases, this data is available to the food manufacturer, perhaps for a price.  The challenge is how to acquire it, surround it with useful context (including your internal data), present it in a way that facilitates understanding, and do so quickly, so that the information still is “fresh.”  Luckily, that’s what BI, supported by data warehousing and other IT techniques and technologies, does very well.

The application of BI that comes immediately to most people’s mind is one that uses internal data, such as from ERP and CRM systems, but those aren’t the only possible applications.  All you need is a reliable, high-quality, automated source of external data, and BI can use it just as well as your internal data.  A BI application that helps you keep your finger on the pulse of the aspects of the current recession that potentially affect your business can enable the intensive monitoring recommended by McKinsey and help your company survive – and perhaps even flourish – in these difficult times.

To see past comments from The Intelligent Food Chain, please click here to enter our Archives Section.