For most mid-market distributors and manufacturers, the tools that work at 200 SKUs stop working somewhere between 1,200 and 1,800 — not because the tools are broken but because the hidden assumptions they were built on stop holding.
The first sign is usually cycle-count variance. At 200 SKUs, a weekly full count takes an hour and variance stays under 1%. At 1,400 SKUs, a weekly full count isn't feasible, cycle counts are inconsistent, and variance creeps into the 3–6% range. That variance compounds: inaccurate counts cause stock-outs, stock-outs cause expedited purchase orders, expedited POs cause landed-cost surprises, and landed-cost surprises cause the CFO to demand the number the floor team can't give.
The second sign is lead-time drift. At low SKU counts, you know your top-10 vendors personally. You feel it when lead times start slipping. At higher SKU counts, lead-time drift is noise — until your 80th-percentile is 3 weeks longer than it was a year ago and nobody noticed. The mean looks fine; the tail is broken.
The third sign is reorder-point staleness. Spreadsheet-based reorder points are set and forgotten. They drift from demand. At 200 SKUs, a quarterly review is enough. At 2,000 SKUs, no human is going to review reorder points quarterly. You need the tool to propose adjustments with a reason and let a buyer accept or override in under five seconds.
The fourth sign is picker training time. This is the one CFOs miss until it hurts. With paper pick lists and a legacy POS, a new picker ramps in 2–3 weeks. With a purpose-built WMS on a Zebra TC21, a new picker ramps in 3–5 days. At factory wages, that difference is ~$3,800 per picker — and your attrition rate determines how often you pay it.
If three of those four signs are showing up, you're at the inflection. This is the moment to consider a purpose-built inventory + warehouse system. We'd like to pitch CloudPeak — but if your volume is under 500 SKUs and your budget is under $10k/year, we'll tell you to keep using Shopify Stocky. We mean that.