How Leaders Balance Supply Chain Inventory Buffers with Cash Discipline

I learned this the hard way in 2019 when tariff threats were swinging wildly and half my e-commerce clients were panic-buying six months of inventory while the other half froze completely. The signal that saved me? Days of inventory on hand relative to your cash conversion cycle.

Here's what I mean. When I was running my fulfillment operation, I had one supplement brand storing 180 days of inventory with us because their manufacturer was in Shenzhen and they were terrified of stockouts. Sounds smart until you realize their cash conversion cycle was 45 days—they collected from customers way faster than they paid suppliers. They were tying up $400K in inventory they wouldn't sell for half a year while their supplier offered 60-day payment terms. That's dead cash earning nothing.

The threshold I used: never carry more than 2x your cash conversion cycle in inventory unless you have data proving that specific SKU will be unavailable or 30% more expensive within that window. Not a hunch. Actual supplier communication or tariff schedule. One of my DTC brands nailed this during COVID—their supplier sent documentation that raw material costs were jumping 40% in eight weeks. They pulled the trigger on a big buy because the math was clear: even if they tied up cash for 90 days, they'd save more on unit cost than they'd pay in opportunity cost or interest.

The brands that got crushed? They carried extra inventory "just in case" with no specific threat and no calculation. When demand shifted, they sat on obsolete stock and had zero cash to pivot into what customers actually wanted. I watched a home goods company burn through their cash cushion on inventory, then couldn't afford to test new products when buying patterns changed. They went under with a warehouse full of stuff nobody wanted anymore.

The right answer isn't always "stay light" or "stock up." It's knowing your cash cycle and demanding concrete evidence before you deviate from it.

I treat the choice as a trade-off between working capital and service level: replenish when projected demand during supplier lead time plus a small safety buffer exceeds on-hand stock; otherwise, conserve cash and avoid tying up capital in slow-moving canvases. Practically, we monitor sell-through trends across replenishment cycles and supplier lead times; two consecutive drops in sell-through paired with longer-than-expected lead times is the signal that tells me to curb orders rather than carry extra inventory.

That rule saved us stock and cash during a period of volatile international shipping, and a broader six-month pricing and forecasting project later reduced unsold inventory by 28%, validating the approach.

We focused on forecast error after removing the most volatile inputs. We removed large promotions, one-time media surges, and unusual referral traffic, then compared the cleaner demand with our plan. When the adjusted forecast stayed above plan for two weekly cycles, we approved a deeper inventory position. This gave us a clearer view of demand that was more likely to last.

That approach proved useful because noisy demand can create false confidence. Clean demand is harder to produce but much more reliable. We found that when adjusted demand stayed above our threshold, the need for more inventory became much more likely. This gave us a better guide than total revenue because it helped us focus on demand that was more likely to continue.

Working across multiple large-scale third-party logistics environments, I watched clients face this exact dilemma repeatedly, especially during periods of supply disruption.

The signal that consistently proved most reliable was velocity consistency—not just how much of a SKU was moving, but how consistently it was moving. A SKU moving 1,000 units a week with low variance is a completely different inventory decision than a SKU moving an average of 1,000 units a week but swinging between 200 and 3,000 depending on the week.

When velocity was consistent, we could recommend leaning toward cash preservation. The supply chain had enough predictability to absorb some lead time risk. When velocity was erratic, carrying buffer stock was almost always the right call, because the cost of a stockout in an unpredictable demand environment far exceeded the carrying cost of the inventory.

The threshold that guided the call was that if the coefficient of variation on weekly demand exceeded 30%, we treated it as a high-uncertainty SKU and recommended buffer. Below that threshold, cash preservation was defensible.

In hindsight, this signal proved right because it was grounded in actual movement data rather than gut feel or blanket policy. The organizations that got into trouble during supply disruptions were almost always the ones applying the same inventory logic across their entire SKU base, regardless of how differently each SKU actually behaved.

My suggestion is: before you decide between carrying inventory and preserving cash, segment your SKUs by demand variability.

In a consumable product with a shelf life, this is not a philosophical question. Stock ages whether or not it sells.

The signal I use is cover measured against lead time rather than against a forecast. Our manufacturing lead time is long, so the floor is lead time plus a month of cover. Below that, I am exposed to an event I cannot respond to. The ceiling is a cash rule: stock never goes above about 40% of available cash, whatever the volume discount looks like. That threshold exists because a discount is only cheap if the money was doing nothing else.

The call that proved right was refusing a bigger run when one good month made the forecast look like a trend. It was a single strong month with a promotion behind it, not new baseline demand, and I could not have funded the following batch if it had sold slowly. We ordered smaller and paid more per unit. Demand did soften. The stock we did not buy would have been sitting in a warehouse with its date printed on the side of it.

The question I would offer anyone is what the stock costs you if it moves at half the speed you expect. If the answer is uncomfortable, that is not a discount. It is a bet with a countdown on the label.

When faced with uncertainty in the supply chain, it comes down to the risk associated with the product when deciding whether to carry extra inventory or remain cash lean. We cover essential, high-volume items with buffer inventory and hold cash on variable design/demand items.

The threshold we operate with:

We monitor Supplier Lead Time Variance. If the actual times are more than 50% off a lead-time norm in two consecutive quarters, we pull from a Just-In-Time system to hold a 60-day safety stock of a critical item.

How this rule paid itself off:

When transit times jumped to double what we anticipated from China due to congestion in ports, our lead time variance triggered the rule even before the situation had escalated. We spent available cash on locking in a buffer of inventory before transit freight went up and the factory ran later. While cash-lean competitors lost shelf space due to out-of-stock items, our supply chain did not falter. When lead times stabilized, we went back to a leaner stock and recovered our cash.

In my experience, the call comes down to where the uncertainty actually sits, so I buffer the items with the worst lead-time variability and keep cash light everywhere else. Supply risk is never spread evenly across a catalogue, and treating it as if it is means you pay to protect lanes that were already stable while the volatile ones still stock out. The signal that guided my call was lead-time variability, the spread between best and worst case on a supplier, not the average lead time. When that spread widened while the average stayed flat, I read it as the lane starting to break and covered it early. Averages looked fine right up to the day they did not, and the spread moved first every time. What I would do is rank suppliers by that spread, hold cover only on the top slice, and let the stable lanes run lean.

There's been many times we've gone through this. I have 10 locations at this time, so I have each location add a few extra to their deliveries and then continue to order normal PAR after that. If the shortage happens, we are covered and can move things between stores if the situation is isolated, or if it doesn't happen, we aren't overstocked and will be able to compensate after the threat is over.