They are figures that make a C-suite officer’s eyes light up. A global fashion retailer increases its market share by more than 28% and doubles its operating profit, all in just three years. It wasn’t because of a new marketing strategy or hot acquisition. Instead, this retailer invested in a serious supply chain digitization strategy.

It’s amazing how quickly balance-sheet improvements occur from supply-chain related changes. Service-level improvements that better the customer experience--as well as reducing lost sales, inventory and waste--are not glamorous initiatives that make headlines. But they work.

I talk to many Chief Supply Chain Officers (CSCOs) and Chief Operating Officers (COOs) who hesitate to go this route because they feel digital supply chain is an all-or-nothing proposition. They envision huge dollar signs and years of work before a payoff.

I’m here to say, “Not so.” Can you go the all-in route? Of course. It works, if you approach it correctly. But the more realistic route, for many companies, involves moderate financial investments in stages that lead to the same end goal. A successful digital transformation can be a series of journeys that lead to a transformed future supply chain. Three priority areas are key to success as you digitize:

  • A unified, single view of demand
  • Supply chain segmentation
  • Smart planning and execution

I could go on for hours about all of this (which doesn’t always make me a hit at parties), but I won’t. Let’s use demand planning as an example of how digital transformation can move your supply chain ahead.

<<< Start >>>

<<< End >>>

<<< Start >>>


A global fashion retailer increases its market share by more than 28% and doubles its operating profit, all in just three years. It wasn’t because of a new marketing strategy or hot acquisition. Instead, this retailer invested in a serious supply chain digitization strategy.


<<< End >>>

A new take on demand planning

Here’s how demand planning has traditionally happened: Different functional areas—operations, finance, trade, and sales—use standard statistical techniques to generate their own forecast using historical sales data and some external data.

Because each functional area “owns” a different forecast, they must come together in consensus meetings to agree on a compromise. The coffee flows, the conversations get more complex as time goes on—you get the picture. As you can imagine (or maybe know from experience), this takes a long time—usually a month or longer. So the sales data is dated before it’s even able to be useful.

If you’ve ever sat in these meetings, you may have found yourself thinking that the process is upside-down. Rather than agreeing on the data and letting the analytics generate a single forecast, the discussion is typically focused on how to find the right balance between conflicting forecasts. Why are these forecasts conflicting? Because they are generated by different functional areas, each of which has a different responsibility and objective. As a result, it is not clear that the consensus forecast correctly represents market behavior.

Let’s look at a more modern approach to demand planning. Say a consumer packaged goods (CPG) company collects data in four main categories that relate to its products and supply chain:

  • Internal data. Historical sales, shipments, prices, SKUs, and more.
  • Consumer data. Point-of-sale information or syndicated data from companies like IRI and Nielsen.
  • Socioeconomic information. Consumer Purchasing Index, unemployment and inflation rates, etc.
  • External data. Google trends, social media product mentions, and more.

Now let’s say they use this data to generate a supply plan, financial plan, and sales plan for the next 50 weeks. This is their planning horizon. Let’s walk through the steps they’ll use to plan for demand.

<<< Start >>>

<<< End >>>

Step 1: Trade plan. The team meshes trade planning information like future promotions and marketing investments with the data described above to generate a demand forecast—by SKU, retailer, and week combination for the entire planning horizon.

Step 2: Demand forecast. The team uses the demand forecast from Step 1, together with past CPG-to-retailer historical shipment data, to generate a forecast of each retailer’s future orders. Again, the level of granularity is down to individual SKUs and weeks for the entire planning horizon.

Step 3: Retailer order forecast. The team converts the shipment forecast to a supply plan that considers available materials—from raw material inventories to manufacturing capacity constraints—and maximizes certain performance measures.

Steps 4 & 5: Supply and financial plans. The team aggregates the retail SKU and weekly forecasts, generating a financial forecast at the brand level for every month of the planning horizon. Then, they compare the financial forecast with the company’s business objectives and trade plan.

<<< Start >>>

<<< End >>>

Why consider a change?

Compared to the consensus forecast, the circular process is simpler. It’s about a single forecast, based on integrated data, generated by advanced analytics—which does a better job of forecasting than humans. Why? Because it can consider so much more data at one time, finding patterns it would take humans months to identify. Executives just need to agree on the data and are freed to focus on their sweet spot--strategy. And this generally takes a week, versus the multiple weeks the traditional method of demand planning took.

It’s a first step toward digitizing your supply chain—but one with results that will likely make you want to take the next step.

See more Supply Chain & Operations insights.

Kris Timmermans

Lead – Supply Chain & Operations

Subscription Center
Subscribe to Business Functions Blog Subscribe to Business Functions Blog