Sales history ≠ market demand
Why manufacturers selling through dealers and distributors need to look beyond historical sales to understand what their markets are really telling them.
Historical sales are an important input for any demand forecast. They show what was ordered, when volumes changed, and where growth or decline appeared.
But in an indirect sales model, sales history often measures movement into the channel, not consumption by the end market.
That distinction matters because dealer orders can move for reasons that have little to do with end-customer demand.
A manufacturer may see sales to its dealers increase and conclude that market demand is strengthening. But dealers may be building inventory ahead of the season, responding to improved availability, ordering before a price increase, or positioning stock based on their own expectations.
The reverse can happen too. Dealer orders may decline while underlying market demand remains healthy because the channel is temporarily reducing inventory.
In both cases, sales history is telling us something important. It just isn’t necessarily telling us what we think it is telling us.
What does your sales history actually measure?
In many B2B industries, demand forecasting is complicated by the gap between sell-in and sell-out: what a company sells into its channel, and what end customers actually buy from that channel.
Consider a manufacturer of construction equipment selling machines through a dealer network across Europe. The manufacturer’s sales data gives a clear view of sell-in: what has moved into its dealer network.
What the planners wants to understand, however, is underlying market demand: how many machines the market is consuming, and how that demand is likely to develop.
Those two signals are connected, but they are not identical. Dealer inventory sits between them, alongside dealer expectations, commercial incentives, backlog, seasonality and local market conditions.
Together, these factors can amplify what is happening in the market. This is one form of the bullwhip effect: small shifts in end-market demand can become amplified as they move upstream through dealer orders, inventory decisions and replenishment behaviour.
Imagine end-market demand softens by 5%. Dealers notice the slowdown and reduce their own inventory as well. Their orders to the manufacturer may fall by considerably more than 5%.
Looking only at sales history could suggest a sharp decline in demand, when part of the drop is actually an inventory correction within the channel.
So what is the market actually telling you?
Historical sales are often not the only source of information available.
Depending on the industry, manufacturers can enrich sales history with market registrations, industry volumes, construction permits, dealer inventory, installed-base information, commodity prices and economic indicators.
Of course, in real life, “available” sometimes means “someone knows where the file is.” The goal is not to find the perfect signal, but to avoid relying on only one.
Commercial teams and dealers can add another layer: upcoming projects, customer behaviour, competitor activity or local market developments. Statistical forecasting adds a more structured view by identifying patterns in historical demand without being influenced by the latest market conversations.
In practice, these signals rarely line up neatly. A statistical forecast may show stable demand while dealers are more positive. Market data may show the industry contracting while your own sales continue to grow.
That disagreement is not just noise, but something to investigate.
Why are dealers more optimistic than the market? Are we gaining share? Is one country behaving differently? Are dealer orders reflecting end-customer demand, or an inventory build?
Those are more useful planning questions than simply asking whether next month’s forecast should be 950 or 1,000 units.
From extrapolating history to understanding demand
The objective isn’t to stop using historical sales. It remains an important demand signal and an essential foundation for statistical forecasting.
But in an indirect sales model, it should be treated as one signal among several.
In real implementations, this starts with visibility: sell-in, sell-out and channel inventory by customer, product and region. Even partial visibility is better than treating the sell-in/sell-out gap as a black box.
The next step is interpretation. What is driving the movement? End-market demand? Timing? Inventory correction? Promotion? Supply constraint? A commercial assumption?
AI-enabled planning capabilities can support this work by comparing signals at scale and highlighting where they diverge. That helps planners spend more time on the exceptions that need judgement, and less time piecing the signals together manually.
In short, historical sales will always have a role to play. They show patterns, seasonality and past behaviour. But they are only one part of the picture, and on their own, they can make it too easy to overlook what is starting to shift.
Want to see this in practice?
Join us on October 21, when I’ll speak with Chantal Esterhuyse from Kobelco Europe about how they combine smart algorithms, dealer input and market intelligence in their demand planning process.
