Retail Inventory Buyers Share How They Hedge Long Lead Times Without Overbuying

Purchasing items in advance and risking stock-outs is really two different questions that have been inaccurately pictured as a strict choice between two extremes, when in fact there is more nuance to the area of inventory velocity than just a binary decision point. We can instead categorize our inventory into different 'Criticality-Velocity Matrix' (or 'Velocity Matrix') segments based on two distinct criteria; what impact does an item's operational use or how fast do we typically 'consume' that item. For high-criticality items (i.e., those needed for continuous operations), we will utilize an alternative approach via use of a Criticality-Velocity approach to develop our reorder point (ROP) based on the most recently available supplier reliability data and not just using a long-term average.

The 'Reliability-Buffer' adjustment approach is my primary guideline when it comes to deciding how much of a buffer to carry for orders being placed with suppliers. If a supplier has shown erratic shipping performance then we will move away from using a static re-ordering approach and will instead rely on their most recent 90 day moving average of actual shipped performance to assign the proper amount of safety buffer for that specific supplier's SKUs. When the shipping variance from the supplier increases, we will subsequently increase the amount of safety buffer that we are carrying for that specific supplier's SKUs. This will allow for only carrying the minimum 'uncertainty cushion' needed to cover the delays in shipping from the supplier. Additionally, this will allow for Replenishment that has been automatically linked to a supplier's most recent shipping performance, keeping cash flow intact and maximizing the Company's agility to change direction if and when needed if the supply chain becomes stable.

Spent years as a plant scheduler and supply chain leader before moving into ops software -- the "buy early or hold cash" tension was a weekly conversation.

The checkpoint I kept coming back to wasn't lead time length, it was schedule risk. If a stockout on that item would trigger a line stop or cascade into missed shipments, it earned buffer stock. If a shortage just meant resequencing jobs, it stayed lean. That one filter cut a lot of the noise out of the conversation.

The harder problem I watched manufacturers miss wasn't the buy decision -- it was that they had no real-time signal telling them when a core item was drifting toward that risk threshold. By the time a spreadsheet or a weekly report surfaced it, the lead time window had already closed. That's exactly the kind of visibility gap we built Thrive's maintenance and parts tracking tools to close -- so the floor isn't chasing data when the decision still matters.

One practical thing that helped on the operational side: stop treating all "critical" parts as one category. We'd separate parts by replaceability speed, not just usage frequency. A part you can source from three local distributors in 48 hours behaves differently than one with a single overseas source -- and they shouldn't sit in the same reorder logic.

My most reliable checkpoint is asking: "Would a customer design change fix this problem?" If the answer is yes, then I avoid early buying because demand is flexible. But if a customer cannot easily switch once a decision is locked, like approved flooring specs in an ongoing build, I secure inventory early. This distinction between flexible and locked demand has helped me prioritize correctly when everything looks urgent on paper.

We focus on splitting SKUs into core and opportunistic categories. For core items, we build in buffer stock based on historical variability and supplier reliability. The rule of thumb: if a supplier's lead time drifts beyond 30 days, we either place an earlier buy or identify an alternative. For everything else, we keep lean and accept some risk of stockout.

We import fragrance from European houses where the official lead time is 4 weeks and the practical reality is 8 to 12 weeks depending on the season. After 9 years of running PerfumeM I've stopped trying to forecast lead time and started forecasting which SKUs deserve the working capital risk.

The decision checkpoint we use: 30 percent customer repeat rate at 90 days.

If a SKU has at least 30 percent of buyers coming back within 90 days, the demand is real. We place early buys against forecast demand even if it stretches working capital. The repeat rate has predictive value that survives supply chain noise. If a SKU is below that threshold, we hold cash. A stockout on an unproven SKU costs us a single transaction. Excess inventory on an unproven SKU costs us 9 to 18 months of dead capital and storage.

The mistake that gave us this rule. In late 2022 we placed an early buy on a niche launch that looked great in our discovery data. Repeat rate at 90 days came in at 11 percent. We were sitting on a six figure inventory position with 11 month sell-through. The cash drag was painful enough that we now treat 30 percent as the floor for any early buy decision.

For core items that already cleared the threshold, the rule of thumb is simpler. We carry 8 weeks of forward cover at current sell-through. If the supplier signals a delay, we pull a buy forward by 4 weeks. If sell-through accelerates above forecast, we pull forward by 6 weeks. The trigger is always sell-through velocity, never the calendar.

The honest bottom line. Most stockout pain in small retail isn't a supplier problem, it's a demand qualification problem. Tighten your repeat-rate signal and the lead time question gets a lot less stressful.

Ahmad Khan, founder of PerfumeM (perfumem.com)

In the watch trade, the $13,000 loss from a bad deal forced me to treat every purchase like a controlled test rather than a timing guess. I now run a quick authenticity and condition check on any incoming piece before committing cash, which tells me fast whether the watch is worth locking in early or better left for consignment.

That same habit keeps inventory lean. I only buy outright when the client-supplied details already match a verified buyer need; otherwise I shift the item to consignment so cash stays free and the watch is still available without excess stock sitting idle.

The checkpoint I use is simple: after the initial valuation call, if the agreed margin is clear and the piece passes the same in-house checks we run on every sale, I buy early. Everything else stays on the client's timeline, which has kept core Rolex and AP references moving without overcommitting.