Demand Planning Case Study: Improving Availability and Freeing Cash

This demand planning case study is a hypothetical example, written to show how the pieces of a planning improvement fit together and how to read the numbers. It is labeled clearly: Illustrative example using hypothetical data. It describes no real client, and the figures are invented for the purpose of explanation. It follows a consumer product company that wanted better availability and less cash tied up in stock, and it traces the diagnosis, the actions, the result and the lessons, with the arithmetic checked. It is part of demand planning and inventory optimization, follows the demand planning audit checklist and the inventory and forecasting health check, and leads to planning transformation and the demand planning maturity assessment. The finance side is in cash flow management.

The situation (hypothetical)

A consumer products company with annual revenue of $80 million sold through retailers and its own website. It carried about 1,800 active items, mostly in four categories. Cost of goods sold was $48 million a year and inventory averaged $14.0 million. Fill rate on retailer orders was 91 percent, below the 95 percent that its main customers expected. At the same time, a growing share of stock was old. The finance director was concerned about cash and a covenant on a credit line, and the operations director was concerned about service. Each blamed the other function's forecast.

The starting metrics were these. Inventory turnover on a cost basis was 48 / 14 = 3.43 times, and days of inventory were 14 / 48 × 365 = 106.5 days. Forecast bias at the item-month level was about +9 percent, meaning forecasts exceeded actual demand on average. Roughly 18 percent of inventory value was older than 180 days.

The diagnosis

A short review found five root causes. First, forecasts were consistently high because the sales team's inputs were adopted without challenge, which produced excess in slow items, and buyers then lowered orders on instinct for others, which produced stockouts. Second, safety stock was set at a fixed four weeks for all items, regardless of how variable demand was or how long supply took. Third, lead times in the system were years old and often too short, so reorder points were triggered late. Fourth, there was no monthly cross-functional meeting; sales, operations and finance talked bilaterally. Fifth, nobody reviewed aged stock until the year-end write-down discussion.

The actions

The company took six steps over twelve months. It segmented items by value and variability. It adopted statistical baseline forecasts for stable items, with a documented process for adjustments and a rule that every manual change must carry a reason. It started to measure forecast accuracy and bias monthly, and discussed bias at the planning meeting. It updated lead times for the top items by value from actual receipts. It reset safety stock by segment, with service targets of 98 percent for the most critical items and lower for the long tail. It set up a monthly S&OP meeting with sales, operations and finance, anchored on one set of numbers, and a quarterly review of aged stock with clearance plans. See safety stock formula, forecast bias and financial reconciliation in S&OP.

The results (hypothetical)

After twelve months in this example, revenue had grown modestly, and cost of goods sold was $50 million. Average inventory had fallen to $11.9 million. Fill rate on retailer orders reached 96 percent. Forecast bias at the item-month level had fallen from +9 percent to about +2 percent. The share of inventory older than 180 days had fallen from 18 to 11 percent.

The calculations are as follows. Inventory fell by 14.0 - 11.9 = $2.1 million, or 2.1 / 14.0 = 15 percent. Turnover rose to 50 / 11.9 = 4.20 times, and days of inventory fell to 11.9 / 50 × 365 = 86.9 days. Check: 11.9 / 50 = 0.238, and 0.238 × 365 = 86.9. The days of inventory fell by 106.5 - 86.9 = about 19.6 days. At a carrying cost of 18 percent, the lower inventory saves 2.1 × 0.18 = $0.378 million a year, and at a financing cost of 9 percent the interest saving alone is 2.1 × 0.09 = $0.189 million. Part of the reduction came from selling down excess, which in this example recovered less than cost on some items and led to markdowns and write-downs. Those costs would need to be set against the benefits. In this example they are assumed to total $0.4 million, a one-off cost. The first-year carrying cost saving of $0.378 million is therefore slightly below the clearance cost, so the first-year net is about -$0.022 million (0.378 - 0.400), and the benefit shows from the second year, when the saving recurs without the clearance cost.

What drove the improvement

Three levers did most of the work in this example. Removing the bias reduced new excess without hurting service, because orders matched demand. Resetting safety stock by segment moved stock from low-value, stable items to critical, variable items, so service improved while the total fell. Correcting lead times prevented late orders. The monthly meeting did not change the numbers directly, but it created accountability for them and a route for resolving disagreements. The company did not buy new software in the first year; it used Excel and its existing ERP with better inputs and routines.

What this example does not prove

This is a constructed example. It does not show what any specific company would achieve, and it does not mean that each of the effects would be of this size elsewhere. Results depend on the starting point, the product mix, the supply base, how quickly the team can change habits and market conditions. Improvements in fill rate and inventory do not always move together; in some cases, achieving a higher service level requires more inventory, and the better outcome is a smarter allocation of it. Use the example to understand the logic and the measures, not to forecast your own result.

Lessons

Start by measuring: bias, accuracy, fill rate, aged stock and days of inventory. Fix the inputs before the tools: data, lead times and parameters. Give the forecast an owner and require reasons for changes. Set service targets by segment. Create one forum where sales, operations and finance work from the same numbers. Plan for the cost of clearing old stock. Track benefits in the finance reports so that the gains are visible and credited.

How to apply this to your company

Calculate your own starting numbers: inventory turnover and days of inventory, fill rate, forecast bias, and the share of stock over 180 days. Identify the two or three biggest causes using the checklist in the demand planning audit. Choose a small set of actions and an owner for each. Set targets and a review date. If you would like an outside view, see the inventory and forecasting health check or our contact page.

FAQs

Is this a real client case?

No. It is an illustrative example using hypothetical data, written to show the logic and the measures.

How long does an improvement like this take?

It depends on the company. In this hypothetical example the changes were made over twelve months, and benefits grew over time.

Do I need new software to get these results?

Not necessarily. In the example, better inputs, routines and policies using existing tools drove most of the gain.

Why might results be different at my company?

Starting point, product mix, supply base, team and market conditions all differ, so results will differ.