Demand Planner interview questions, with worked answers

How to use these questions

Interviewers in planning often show a chart of sales and forecast and ask what you see. Look for patterns before numbers: a trend, a season, a break, a forecast always on one side. Then give a real range you planned, the change you made and the effect on accuracy, stock or service.

What a demand planner is expected to know

• The method: statistical forecastingChoose and tune statistical forecast models, and know when to trust them and when to override., its models and its error measures.
• The process: demand planningOwn the forecast of a product range: build it, check it with sales and keep it honest every month. each month and S&OPPrepare the demand side of the monthly S&OP cycle and bring the open questions to it. around it.
• The link to supply: safety stockSet safety stocks and targets that balance service to customers against the cost of stock., service levels and MRPKnow how the forecast drives material requirements planning and what happens when it is wrong..
• The data: ExcelClean data, check a model's output and build quick analyses when the planning tool cannot., SAPRead and maintain sales history, material data and plans in SAP. and the planning system.

Planning tools are named in ads side by side and none is a must: SAP IBP, Kinaxis, o9 and Blue Yonder are common systems, and Excel sits next to all of them. Interviewers care more about how you use a model and question its output than about the product.

The forecast and its errors

How do you build a baseline forecast for a product?

What they check: the starting point of every plan.

Short answer: "A baseline forecast is what we expect to sell without special events such as promotions. I clean the history first, removing one-off orders and weeks with stock-outs, then let the statistical forecast pick up level, trend and season, and check the result against what I know about the product. Promotions and events are added on top, separately."

The full answer: A clean history matters more than the choice of model; most bad forecasts start with bad data.

A weak answer: "I take last year's sales." Follow-up: "Last year had a big one-off order." Say what you do with it.

How do you measure forecast accuracy, and which measure do you report?

What they check: measuring the right thing.

Short answer: "Forecast accuracy compares the forecast with what really sold, usually through an error measure such as the mean absolute percentage error or a weighted version that gives large products more weight. I measure at the level and the time lag where decisions are made, such as product by month three months ahead for production, and I always report bias next to it."

The full answer: An accuracy figure without its level and lag says little; state both.

A weak answer: "Our accuracy is good overall." Follow-up: "Accuracy is high in total but plants complain." Say what you look at.

What is forecast bias, and how do you fix it?

What they check: honesty in the numbers.

Short answer: "Forecast bias is a forecast that is wrong in the same direction again and again, usually too high because sales want stock to be safe, or too low to beat targets. I track it per product group and per person who adjusts the forecast, show it in the demand review, and agree with sales which adjustments help and which only add bias."

The full answer: A small error that changes sides is normal; an error always on one side is a habit to break.

A weak answer: "Some months high, some low, it evens out." Follow-up: "Sales raise every forecast by a margin of their own." Say what you show them.

How do you handle promotions in the forecast?

What they check: the events that break simple models.

Short answer: "I keep promotions apart from the baseline: I get the plan from marketing or key accounts early, estimate the extra volume from past promotions of the same kind, and include the dip before and after and the effect on similar products. After each promotion I compare forecast and sales and keep the lesson for the next one."

The full answer: The promotion you hear about late is the one that empties the shelves; build the habit of early notice.

A weak answer: "Sales tell me the number." Follow-up: "A promotion sold twice the forecast." Say what you check first.

How do you forecast a new product with no history?

What they check: judgement where models cannot help.

Short answer: "I look for a similar product launched before, a like-for-like, and adjust its curve for price, channel and marketing. I agree the launch volume with marketing and sales, plan for the old product it replaces, and review the forecast weekly in the first months, because the first weeks of sales say a lot."

The full answer: Plan the launch as a range of outcomes and agree with supply what to do if sales come in high or low.

A weak answer: "Marketing's number goes into the system." Follow-up: "The new product sells half the plan after a month." Say what you change.

How do you deal with seasonality and outliers in the history?

What they check: reading sales data well.

Short answer: "Seasonality is the pattern that repeats each year, such as summer peaks; the model needs a few years of clean history to learn it. An outlier is a value far from the pattern, such as a one-off order or a stock-out. I find outliers with simple rules and a look at the chart, correct or remove them, and note why, so the next planner understands the history."

The full answer: Never delete an outlier you cannot explain; ask sales first, it may be a new customer.

A weak answer: "The system handles outliers." Follow-up: "A pandemic year sits in your history." Say what you do with it.

S&OP, supply and stock

Walk us through the S&OP process and the demand planner's part in it.

What they check: S&OPPrepare the demand side of the monthly S&OP cycle and bring the open questions to it. from the inside.

Short answer: "S&OP is the monthly process in which a company agrees one plan for demand, supply and finance: a demand review, a supply review, a meeting to settle gaps and an executive meeting that decides. The demand planner owns the demand review: the agreed forecast, its risks and the scenarios, such as a large customer won or lost."

The full answer: The executive meeting should decide, not discuss numbers; prepare the decisions in the steps before.

A weak answer: "S&OP is a monthly meeting." Follow-up: "Executives change the forecast in the meeting." Say how you respond.

How do you run a consensus forecast meeting with sales?

What they check: collaboration that improves the number.

Short answer: "A consensus forecast is the one number that sales, marketing and planning agree on. I send the statistical forecast and the big changes before the meeting, ask each team for the reason behind any change they want, and record who changed what, so we can see later whose input made the forecast better. The meeting is short and covers only the exceptions."

The full answer: Ask for reasons, not numbers; a reason can be checked, a number only argued.

A weak answer: "We average everyone's numbers." Follow-up: "Sales skip the meeting." Say what you do.

How does the demand plan become a supply plan?

What they check: knowing what happens downstream.

Short answer: "The agreed forecast goes into supply planning, which checks it against stock, capacity and suppliers and plans production and purchases; MRPKnow how the forecast drives material requirements planning and what happens when it is wrong. then works out the materials needed. Where supply cannot meet demand, the gap comes back to the S&OP process to be decided. So a wrong forecast becomes wrong orders all the way to the suppliers."

The full answer: Know the lead times of your products; a change inside the frozen period costs far more than one outside it.

A weak answer: "Supply is another team's job." Follow-up: "The plant cannot make the forecast in March." Say what happens next.

How do you set safety stock for a product?

What they check: inventorySet safety stocks and targets that balance service to customers against the cost of stock. with numbers behind it.

Short answer: "Safety stock is the extra stock kept to cover forecast errors and delays in supply. I set it from the size of the forecast error, the supply lead time and its variation, and the service level the business wants for that product, which is the share of demand we aim to deliver from stock. Important products get a higher service level, slow ones a lower one."

The full answer: Better forecasts lower the safety stock needed; show the business the stock that accuracy frees.

A weak answer: "Two weeks of sales for everything." Follow-up: "Finance wants a lower stock value." Say what you offer.

Stock is too high and service is too low at the same time. What do you look at?

What they check: diagnosing a classic problem.

Short answer: "Usually the stock sits in the wrong products or places: too much of slow products, too little of the fast ones. I split stock and service by product, find where the forecast was biased, check safety stock settings and order sizes, and look at the products no longer sold. Then I fix the worst groups first with the supply team."

The full answer: Total stock hides the problem; it is always in the detail.

A weak answer: "Cut stock everywhere." Follow-up: "A large customer complains about missing deliveries." Say where you start.

How do you decide which products get most of your attention?

What they check: focus in a range of thousands of items.

Short answer: "I group SKUs, the single stock items, by their value and by how predictable their demand is. Large, stable products get a good model and a light check; large, unstable ones get most of my time and the sales team's input; small, stable ones run on the model alone; small, unstable ones may need a different supply approach rather than a better forecast."

The full answer: Tell the business which products you will not forecast in detail, and why.

A weak answer: "I check every product each month." Follow-up: "The range grows by a third." Say what you change.

Tools, sensing and AI

What is demand sensing, and when does it help?

What they check: knowing the newer methods and their limits.

Short answer: "Demand sensing uses the most recent data, such as daily orders, sales at the till or shipments, to correct the forecast for the next days and weeks. It helps fast-moving products with good daily data, such as consumer goods. It does little for long lead times, where decisions are made months ahead."

The full answer: Sensing improves the short term; it does not replace a good monthly process.

A weak answer: "It makes forecasting obsolete." Follow-up: "The plant needs the plan three months ahead." Say what sensing changes for it.

How do you choose and use a planning tool?

What they check: planning toolsWork in a planning system day to day: models, exceptions, scenarios and the master data behind them. used well.

Short answer: "I start from our process and data: how many items, how many levels, how often we plan, which systems feed it. Tools such as SAP IBP, Kinaxis, o9 and Blue Yonder differ in how they model supply, scenarios and collaboration; Excel stays useful for checks and quick analyses. Whatever the tool, master data and a clear process decide most of the result."

The full answer: A new tool on old habits gives old forecasts faster; change the process with it.

A weak answer: "The newest tool is always the right one." Follow-up: "Planners keep working in spreadsheets after go-live." Say why and what you do.

How can AI improve demand planning, and where are the limits?

What they check: today's tools with judgement.

Short answer: "Machine learning models can use more drivers than classic models, such as weather, prices and promotions together, and often help with large ranges. AI can also find outliers and explain changes in plain words. The limits are data and trust: models need clean, long history, and planners and sales must understand why a forecast moved before they accept it."

The full answer: Compare any new model with the current forecast on the same history before you switch.

A weak answer: "AI will forecast everything better." Follow-up: "The AI forecast is more accurate but sales do not trust it." Say what you do.

Senior rounds

As a senior demand planner, how would you coach a junior demand planner?

What they check: building the planning team.

Short answer: "I give a junior demand planner a product group of their own, sit with them in the first demand reviews, review their adjustments and their bias each month and explain my own reasoning on hard products. I let them present a part of the review early, because speaking up to sales is the skill that takes longest."

The full answer: A senior demand planner or demand planning manager is judged by planners who can defend a forecast on their own.

A weak answer: "Juniors learn the system first, then we talk." Follow-up: "Your junior accepts every change sales asks for." Say what you do.

Your CV for these questions

Each job on your CV invites the questions above: which range, which method, which tool and what changed. Give those stories a line each, written so a planning manager wants to hear more:

Put your CV beside a demand planner job you want with the free score; the area it finds thin is the one to rehearse first, with a range you can describe from history to forecast to stock.