How to turn a demand forecast into an inventory plan.
This tutorial follows the calculation from expected demand to a purchasing recommendation, then shows how long-range and short-range planning work together as better information arrives.
- Written by
- Ronnel Boettcher
- Published
- September 1, 2026
- For
- Owners, supply chain leaders, buyers, and planners
Plan the decision, not just the forecast.
A demand forecast estimates what customers may buy. It does not know, by itself, what is already on the shelf, what is on the way, how long a supplier takes, or whether the team is planning a purchase, a production run, or a store delivery.
A complete planning system adds that missing information. It turns an estimate into a recommendation, brings the recommendation to the right person, and records what happened afterward.
That is why forecast accuracy is only one measure of success. The larger question is whether the company can keep products available without holding more stock than it needs or wasting planners’ time.
Start with expected demand. Finish with a feasible order.
Suppose an imported product has a 12-week replenishment lead time. The buyer needs enough supply to cover demand until the next order can arrive, plus an agreed ending buffer.
Forecast demand
1,200
units through the horizon
Usable stock
500
units available now
Confirmed inbound
300
units arriving in time
Target ending stock
200
units held as protection
net requirement = forecast demand + target ending stock − usable stock − confirmed inbound
1,200 + 200 − 500 − 300 = 600 units
The buyer then applies case sizes, minimum orders, supplier capacity, shelf life, storage limits, and approvals. The target ending stock should reflect uncertainty and the desired availability—not a fixed percentage added to every product.
The forecast creates purchasing options.
If an imported source offers roughly 20% lower landed cost but requires 12 weeks, the business can only use it when the need is visible early enough. A reliable long-range plan can unlock that option before the only remaining choice is a faster, more expensive source.
Short-range planning keeps the commitment useful.
Once supply is committed, weekly planning uses fresher sales, stock, and incoming-order data to allocate it, transfer it, or make a targeted top-up. Efficiency and responsiveness work in tandem rather than as separate plans.
From your data to a plan you can measure.
Each step has a clear purpose. When the result is measured, it becomes better information for the next planning cycle.
- 01Reliable data
- 02Demand forecast
- 03Inventory position
- 04Recommended plan
- 05Planner review
- 06Measured result
- 01Reliable data
- 02Demand forecast
- 03Inventory position
- 04Recommended plan
- 05Planner review
- 06Measured result
Where ToolPlex fits
ToolPlex connects the systems behind the plan and gives teams one place to review forecasts, stock, recommendations, and results. Planners can investigate a number, apply their judgment, and see how the decision performed later.
01 / 03
Demand forecast
How we build a planning system.
- 01
Start with the decision
Decide what the system must help your team do: buy, produce, replenish, allocate, or transfer stock. Define how far ahead that decision is made and what a useful recommendation looks like.
A clear job for the forecast and a result the business can measure.
- 02
Make the planning data reliable
Bring together the sales history, stock on hand, incoming orders, product information, lead times, and other inputs needed for that decision. Settle the definitions and add checks before relying on the numbers.
One current view of demand, inventory, and existing commitments.
- 03
Measure the current method
Test the spreadsheet, ERP forecast, last-year rule, or planner method the company uses today. Compare it at the same product, location, and time level where the decision is actually made.
An honest baseline and a fair test for anything new.
- 04
Forecast demand
Compare suitable forecasting methods and keep the simplest one that performs reliably. Seasonality, promotions, new products, periods without stock, and slow sellers may need different treatment.
A practical estimate of what customers are likely to need.
- 05
Account for stock and incoming orders
A demand forecast is not an order quantity. Subtract usable stock and goods already on the way, then apply lead times, case sizes, availability goals, and other real purchasing or production limits.
Recommendations that reflect the company’s actual inventory position.
- 06
Bring the recommendation to the planner
Show the items that need attention, explain why they were flagged, and let planners review or adjust the recommendation. The result should fit the approval and ordering process the team already follows.
A focused worklist instead of another large report to inspect.
- 07
Measure what happened
Track sales, availability, excess stock, planner changes, and the outcome of each decision. Use that evidence to improve the data, forecast, and planning rules for the next cycle.
A planning system that gets better as the company uses it.
Different decisions happen on different clocks.
Supply-chain planning distinguishes decisions by horizon: longer commitments shape capacity and sourcing, while nearer decisions use fresher information to replenish and allocate. They are not separate plans. Each shorter cycle updates how the longer commitment is used.
Purchasing or production
How much will the business need in total?
Often weeks or months ahead
Expected demand, lead times, supplier or factory commitments, and available inventory
Allocation or replenishment
Where should the available stock go?
Closer to the selling period
Current stock by location, recent demand, incoming transfers, and local availability
Exceptions and transfers
What changed, and where should the team intervene?
Daily or weekly
Stockouts, slow stock, unusual demand, late orders, and planner judgment
One inventory rule will not fit every product.
Applying the same stock target to every product and location looks consistent, but it can create excess in slow lines and leave unpredictable items exposed. The first question is not always “how many?”
Fast and regular sellers
How much should we keep available, and how often should we replenish it?
Seasonal or irregular sellers
When is demand likely to arrive, and how much uncertainty should we allow for?
New products
What assumptions and comparable products should guide the first plan?
Very slow sellers
Should this product be stocked at this location at all?
Show people what needs attention.
A planner should not have to review every product and location. Most of the catalogue will not need intervention on a given day. The system should rank the exceptions: likely stockouts, excess stock, unusual changes, late orders, and recommendations that deserve judgment.
The full detail should remain available. But the daily experience should begin with a manageable worklist, the reason each item appears, and the information needed to make a decision.
Use the right method for each part of the job.
A dependable planning system usually combines several kinds of technology. ToolPlex brings them together behind one experience for the people making the decisions.
Data connections
Keep sales, inventory, orders, and product information current without rebuilding the same spreadsheet each cycle.
Forecasting
Use statistical methods or machine learning to estimate future demand and compare the result with the current method.
Purchasing and inventory rules
Apply stock, lead-time, availability, case-size, and business limits to turn expected demand into a practical recommendation.
AI assistance
Let planners ask questions in plain language, investigate unusual recommendations, and find the supporting information without opening several systems.
The same method, shaped by what the product does.
The seven stages above do not change between industries. What changes is the clock the plan runs on, and what counts as stock you can actually sell.
Fresh food distribution
Weekly cycle · warehouse buy plan
Perishable goods change the shape of the same system. The planning cycle runs weekly rather than monthly, because a product with a five-day shelf life cannot be planned a month at a time. Stock already in the warehouse only counts as coverage for the demand it can still reach before it expires, so the purchasing recommendation has to account for shelf life rather than netting a whole horizon against today's position. And waste sits alongside availability as an outcome worth measuring: ordering to a higher service level on a three-day product buys cover with spoilage, which is a trade-off a planner should be able to see priced before committing to it.
Start with one planning decision.
Choose a purchase, production, allocation, or replenishment decision where better information could reduce excess, prevent shortages, or save planners time. Test it with your own history, then build from the result.