Buffer stock isn't optional — it's insurance.

Most e-commerce operators fill in the buffer stock field once, when their inventory system first asks for it, and never touch it again. Half the time it gets left on zero. That's a mistake. Buffer stock is the cheapest protection you can buy against a stock-out, and most operators run their whole catalog without it.

What buffer stock actually is

Buffer stock, sometimes called safety stock at a per-SKU level, is the minimum quantity you keep on hand at all times to absorb variance you can't predict. People confuse it with the reorder point. The reorder point says "order more when you cross this." The buffer says "don't ever go below this; treat it as out of stock."

A SKU with buffer = 1 and available = 1 is, operationally, out of stock. Don't sell that last unit. If anything goes wrong with it (a damaged return, a miscount, a unit that walks), you're at zero with no time to recover.

Why teams skip it

The usual reasons don't hold up.

1. "We'll just reorder faster"

OK, but your supplier still needs lead time. If you're at zero on a Tuesday and your lead time is 14 days, you're at zero through to the next Tuesday, regardless of how fast you act.

2. "We don't want capital tied up"

A buffer of 1–3 units per SKU is rounding error in working capital. Being out of stock costs you the lost sale, the customer who tries a competitor and stays there, and the ad spend you paid to send them to a sold-out page. That dwarfs the cost of holding a couple of extra units.

3. "Our system doesn't really support it"

True for most. Cin7 Core has a min-stock field but no logic that acts on it. The threshold is just for an alert, not for the reorder calculation.

How to set it (without overthinking)

The right buffer for a SKU is the variance you can't predict. Work it from these:

• Lead-time variance. If your supplier hits ±5 days on a 30-day lead, set buffer to cover ~5 days of demand for that SKU.
• Demand spikes. If a SKU has occasional surges (a TikTok mention, a corporate order), buffer should absorb a typical surge size.
• Quality risk. If a small percentage of units arrive damaged or unsellable, buffer covers that loss.

Across our own 800-SKU Cin7 catalog, most stable lines sit on a global buffer of 1. That's the floor. The high-velocity, high-volatility lines get overridden to 3–5, and a handful of seasonal hero products go to 10 or more through peak.

What it changes operationally

The moment your inventory system treats available <= buffer as effectively out of stock, the numbers start telling the truth:

1. Reorder triggers earlier. You start the reorder process before you've actually run out, with breathing room.
2. Out-of-stock dashboards stop lying. A SKU at 1 unit with buffer 1 shows as OOS, which is the truth. You've got one in the warehouse, not "in stock."
3. OOS-since dates become accurate. You can answer "how long has this SKU been out?" honestly, which lets you measure actual fill rate.

In Stocura, every SKU has a buffer field with a global default. The reorder engine, OOS detection, OOS-since stamping, and dashboard alerts all respect it. A SKU with available = 1 and buffer = 1 counts as out of stock everywhere it appears. One definition, applied across every page that touches stock.

Bottom line

Pull your top 20 SKUs by revenue and set a buffer on each one this week. The math takes an afternoon. You'll only notice it the first time a supplier slips and you don't go to zero.