Why businesses are shying away from using Statistical Safety Stock to drive Strategic Inventory decisions?

In my experience, working with many of our clients, I have come to conclusion that systemic method driving inventory decisions will be successful so long the leadership team “is engaged” and sees the benefit of it. It is important to remind ourselves that fundamental change in the ways of working requires cross-functional ownership and an executive champion.

For most companies statistical safety stock calculations are big black box and consultants have not done a great job of explaining what statistical safety stock means and the benefit of moving towards a statistical safety stock versus manual or days forward coverage, which is predominantly used in the industry. If the concept is not understood, leadership sees this as adding on to the business risk and the initiatives get dropped. In many cases inertia of change by itself derails the initiative, planners do not trust the statistical safety stock and continue with their existing methodology for inventory planning.

Success of moving towards an objective and mathematical stocking strategy starts from segregating the business and laying out process for each of the segmentations. Many times safety stock calculations for some products are way off the mark and if no operational override are built into the process, may trigger service issues, which may lead to distrust in the safety stock recommendations and in many cases discontinuation of initiative altogether.

Why Statistical Safety Stock is not used?

Why the preference for Manual or Days Forward Coverage?

1. Manual (min/max/fixed qty) --- Provides more control to planners, inertial of change etc.

2. Days Forward Coverage ---Easy to understand and most supply chain systems software have inbuilt capability to model safety stock based on days of coverage. I think this is good second step and prepares the business to move towards statistical safety stock.

3. Statistical Safety Stock ---Difficult to understand, difficult to acquire all relevant data like historical forecast error, order size, lead-time, bias etc. Most important of all, sufficient resources and training is not provided to project implementation team and business planning team. Poor job is done of communicating the benefits of moving toward mathematical and objective stocking strategy. In most Statistical Safety stock calculations Demand Error is considered, however in many business Supply Error is as big an issue if not bigger. Statistical Safety Stock is presented as full and final, rather than the input to broader stocking strategy, without the option of user override.

Keep in mind, markets are more competitive than ever before. Your product portfolio now has higher percentage of promoted products and/or displays than any time in the history of retail marketing. Additionally you are constantly changing the configuration of your special packs. All things being equal, simple act of increasing promotions or displays should trigger increased inventory in your supply chain network. Shorter customer order lead-times and shorter product life cycles are also adding volatility which is forcing business to keep higher inventory levels.

On the positive side, some retailers are pushing for smaller case size i.e. less order consolidation, CPFR and ECR is helping reduce volatility in the Turn (Non Promoted) business.

Mantra for successfully moving your organization towards Statistical Safety Stock.

1. Get Management buy-in: Establish clear goals, with minimum customer service targets and potential inventory reduction.

• You will need aggregate or roll-up view of data when talking to the management.
• Remember to always communicate that Statistical Safety stock is derived from “assumption of expected error, which is based on historical error”
• If the business environment changes dramatically, due to competition, new product launches, cannibalization, loss of customers, significant change in customer order pattern etc than statistical safety stock will not solve the situation and broader business wide initiative is required to solve the issue.
• Remember leadership is only concerned with the risk and benefit, and will like to mitigate risk. Many time initiatives fails when Statistical Safety is used in planning without the identified risk, business environment changes and Statistical Safety stock is blamed. Remember to communicate that it will be learning process and treat this effort as ongoing process improvement.

2. New Products: Cannot be planned based on Statistical Safety Stock. New Products should be planned based on what I would call “flexible manufacturing strategy working closely with your customers and suppliers”. Inventory decisions during launch and ramp up phase should be guided by historical new item launch data if available. We have implemented a strategy where you segregate inventory requirements between pipeline inventories and turn inventory, it is especially helpful if you are in Foods or Consumer Packaged Goods industry to maintain retailer buy-in matrix.

3. Segregate the SKU portfolio: By sales volumes and order volatility. This is required so different strategies can be applied to different segments

4. Conference Room Pilot: This is where you prove the value of this initiative, so it is important to select a subset of item, where value can be proven conclusively.

• Select a product line which is high in sales volume and low in order volatility. This product line will have least Forecast Error and possibly least supply error.
• Product is probably made at multiple sites and shipment lead-time to customer/retailers is at the lower than the average shipment lead-time. There will be lesser tendency to consolidate orders by your logistics team.
• Statistical Safety Stock calculations will be more stable, you can expect normal distribution, easily build the confidence interval and based on your agreement with management on service levels, you will be able to calculate safety stock in qty or days of coverage.
• Once safety stock is calculated, it is important to compare this with the existing policy. Understand why the existing policy is different than what you came up with. It is important to discuss this discrepancy with wider business team. Many times there are historical business reasons for higher or lower safety stock targets used by business teams. It is important to understand and document all this information.
• While presenting to management, it is helpful to calculate average inventory reduction by month and dollarize the savings. I like using 25% as guideline for calculating net cash saving based on the inventory reduction. E.g. If total projected reduction in inventory is $100 K, then net cash saving will be approx. $25 K. Remember there is always a hidden cost of carrying inventory like warehouse space, insurance, cost of capital, transportation, damage etc. At this time it is important, as earlier, to communicate to management that, Statistical Safety Stock is derivative of “expected error, which is based on historical error” and significant variations in business should be managed outside of this process.
• Apply the new safety stock stocking strategy in your planning process; if you are using more sophisticated planning system, it becomes easier to incorporate this.
• It is important to train the users, as there is tendency to override the new safety stock and get back to old ways of working. You will not see any benefit and will not be able to actualize the savings.
• Measure the savings for at least one quarter (three months) and then present the learnings to broader team, including the management.
• It is important to publish the success when achieved and reiteration by management towards this initiative.
• Continuous monitoring and monthly/quarterly update to statistical safety stock will be necessary.

5. After this you should move to next big segmentation of your portfolio of items and so on.

6. Plant attainment i.e. schedule vs. production should be measured and communicated. Operation overrides should be applied to account for supply error. Changes in market conditions should also trigger some overrides and you will need process for that.

I will summarize by saying that I have seen at least at our clients a shift towards using statistical safety stock in making strategic inventory decisions. However moving towards using mathematical and objective approach to inventory management is a tedious task for both implementation team and the business.