Sales · Operations
When sales and production pull in different directions

Most companies are organised into functions, and each function is given its own targets. Sales has an order-booking target. Production has an efficiency target. Finance has an inventory target. Each function does its best to meet its number.
What happens when the actions of one function become problems for another? Let me explain with a story that will feel familiar to anyone in an engineered-products business.
A long sales cycle meets a quarterly target
Picture a company that makes large, custom-built machines, each worth tens of millions of rupees. Every sale starts with an engineering study of the customer’s needs. After weeks or months of discussion, the customer may buy, postpone or walk away.
A handful of application engineers support a large sales team in these studies. The same engineers also help the plant with technical issues during manufacturing. With dozens of open deals each, the engineers keep switching between them. Every deal moves slowly.
Now add one more ingredient: quarterly targets. The sales cycle is longer than a quarter.
The pipeline arithmetic
Let us do some simple arithmetic.
Suppose 1 in 10 deals is won and the target is 10 wins per quarter.
- If a deal takes one quarter to close, the team needs about 100 open deals at any time.
- If a deal takes two quarters, the team needs about 200 open deals to produce the same 10 wins per quarter.
But it gets worse. With 200 deals open, each deal gets less attention. Engineers switch more often. Deals take even longer to close, and more of them are lost to competitors. So the team needs even more open deals, which stretches the cycle further.
Do you see the loop? A longer sales cycle needs a bigger pipeline. A bigger pipeline makes the cycle longer and the win rate lower.
The “easy” way out: quarter-end discounts
To break this loop, the company does what many do: it offers discounts at quarter end to close deals faster.
At first, it works. Deals close, the quarter is saved. But customers learn quickly. Soon, they simply wait for the quarter end to negotiate. The discount becomes a ritual, and margins shrink every quarter.
Production pays the price next
Now look at the plant. Assembling a machine takes much less time than selling it, so on average the plant has plenty of capacity. But the discounts mean orders arrive in a flood at quarter end and a trickle at other times.
- During the flood, the plant struggles, hires extra skilled people and runs overtime.
- In between, those people sit idle, and the plant’s efficiency numbers look terrible.
Under pressure from finance and HR, production asks sales to give it orders earlier. Sales responds with a clever fix: it releases “likely” deals to production before they are signed. These advance orders go by many names in different industries; the effect is the same.
And sales pays the price again
The plant builds machines for deals that aren’t closed yet. Now finished stock piles up, and there is pressure to ship it by quarter end. The salesperson negotiating the final terms knows the machine is already built, and so, often, does the customer. Who do you think has the stronger hand in that negotiation?
A few more points of discount are given away to close the deal. Sometimes the customer changes the specification at the last moment, and the machine already built has to be reworked.
Everyone is a victim, and everyone is a cause
Let us trace the chain:
- Top management sets quarterly targets that don’t match the sales cycle.
- Sales responds with quarter-end discounts.
- The discounts create spikes for production.
- Production responds by building ahead of signed orders.
- Built-ahead machines weaken sales in negotiation and create rework and finished-goods inventory.
- Margins fall, inventory rises, and the company is surprised that profit doesn’t grow with sales.
Every function acted sensibly to meet its own target. Together, they made the whole company worse off.
When each function optimises its own number, the company pays for all of them.
What would help instead?
The way out starts with seeing the system, not the functions:
Treat the engineers as the constraint of the sales process. They are scarce and shared. Limit the number of complex deals they work on at once, and sequence deals by priority instead of letting every salesperson pull them in all directions. Deals will close faster, and win rates will rise.
The numbers make the point. Suppose each engineer can complete about 3 engineering studies a month, and each study needs roughly a week of focused work. With 30 deals open per engineer, a study waits its turn among the others and takes around ten months to finish (30 ÷ 3). With 6 open per engineer, the same engineer still completes 3 a month, but each study now takes about two months. The customer hears back in weeks instead of quarters, long before a competitor does.
Match the pipeline to capacity, not to a target. A smaller, better-served pipeline can produce more wins than a huge, starved one.
Smooth orders, don’t spike them. Drop the quarter-end discount habit gradually, and give customers a reason to order early (a reliable, short delivery date, for example) instead of a reason to wait.
Build only to signed orders for customised products, and use the plant’s spare capacity in quiet periods for standard sub-assemblies that are needed in every machine.
The deeper lesson is the one Goldratt kept repeating: a system is not the sum of its parts. Improvement starts when sales, production and finance agree on one goal and one set of measures for the whole.