N-Tier Demand Management


Abstract needed here…


The classic bull-whip effect means that the further a supplier is removed from the end-consumer, the worse are the fluctuations in demand that they see. This has led many to recommend an n-tier approach to demand management, where everyone gets visibility to the end-customer demand at the same time. In practice, very few companies have been able to actually realize this vision. There are some practical approaches that a supplier deep in the supply chain can do to mitigate the bull-whip effect.


Outsourcing and leaner supply chains are pushing companies to using networked models (real-time sharing of information across multiple tiers) for demand management. Most networks today are still working on building connectivity with their immediate trading partners, but the real promise comes from connecting n-tiers. Successfully managing an n-tier networked model involves sharing of real-time data such as POS data, orders, and changes in plan. This vision does not necessarily mean that companies need to create a detailed model of the n-tier supply chain and run optimization logic as they might within their own organization. Major practical challenges that have kept many companies from realizing the n-tier vision must be dealt with, specifically:

  • Understanding multi-channel demand
  • Avoiding multi-counting

Managing N-Tier Demand through Multiple Channels

Managing N-Tier Demand through Multiple Channels

Figure 1 – Multiple Channels for Single Customer

In an n-tier supply chain, demand to the supplier from a single customer, such as an OEM, may travel through multiple channels (see Figure 1). It is up to suppliers to aggregate expected demand for each major [1] customer regardless of the channel that it flows through. This requires understanding the total market size, demand elasticity, and share for each customer (see section below “Forecasting Your Customer’s Demand”). In addition, many suppliers fail to effectively aggregate demand by channel for each major customer, although it is critical to understanding overall demand.
Effectively synthesizing demand across multiple channels requires close dialog with the OEM about demand at the finished goods level, and systems that are able to explode a complex demand BOM against OEM finished goods demand (see sidebar). That is why it is important to monitor and forecast OEM sell-through rates[2]. Once these are in place, the dialogs with OEM about demand can shift from discussion about supplier’s parts to discussion about OEM’s finished goods, for which the OEM has longer-term data and forecasts than they have in their MRP plans for individual parts.

Avoiding the Multi-Counting Trap

Suppliers that are several layers removed from the end-customer in multi-tier supply chains are prone to confusion about true end market demand. Multi-counting of the same demand is a common symptom. For example: a large telecommunications carrier replacing their line of cell phones sends out RFQs to multiple phone OEMs, who in turn send RFQs to several contract manufacturers. By the time the demand signal gets to the supplier, it can be grossly overstated.

Figure 2 – Multi-counted Demand Signals

If the supplier is several steps removed from the end-customer, they need to use their own intelligence to ferret out big end-customer deals, to get a more accurate picture of actual demand. As they pursue opportunities with OEMs, the account management team should capture information on potential end-customer deals (e.g. end-customer name, project name, size and type of deal, etc.). This data is factored into the forecasting scrubbing process to eliminate duplicate demand (see sidebar “Rationalizing Demand Across Channels“). Getting salespeople to consistently enter this kind of data is not easy. It must be made nearly effortless, and part of their compensation should be based on consistency and accuracy in tracking large end-customer deals.

Forecasting Your Customer’s Demand

Beyond this, you should forecast demand for your customer’s whole market and their share. For example, if you are a supplier to Ford, you would create your own forecast of demand for the whole light truck market and for Ford’s share. Or a supplier to Juniper Network would forecast the whole enterprise switch market and Juniper Network’s share. This way, if several major customers have aggressive forecasts, you can make your own informed opinions about whether the total market is really growing, or whether there is some double-counting going on. Formulating your own market assumptions provides the checks and balances needed to get to the best number you can for each account.

Taming the Bull Whip

Suppliers that are several layers deep in the supply chain can stop being at the mercy of late or incomplete demand information by:

  • Developing a Demand BOM approach to understanding the actual demand as it flows through various channels
  • Having a disciplined approach to keeping their “ear to the ground” on what is happening at each of the largest downstream channels and end-customers or large OEMs
  • Maintaining their own perspective, getting a good handle on the total market size and using it to sanity check forecasts they receive from their customers

These steps can help upstream manufacturers avoid much of the misery and destructive force of the “bull-whip effect” that is normally the bane of their existence.

[1] Aggregating by customer is only worth effort for your top customers (e.g. the top 20% of customers comprising 80% of demand). Demand from smaller customers can be aggregated by channel.

[2] Sell-through means knowing your channels’ sales data, in this case the actual sales of the OEM, rather than consumption into their production lines.

Scroll to Top