The Data Center Boom Could Make Your Business… Or Break It
Every HVAC distributor has heard some version of the pitch: your ERP handles seasonality, your reorder points adjust, the system “learns” your season. It’s on the industry page of every major ERP vendor, including the one you already run.
Ask a purchasing manager how that’s actually going in June, when the compressor SKUs turn over three times faster than they did in February, and you’ll get a different answer.
The math your system is actually running
Most distribution ERPs, P21 included, calculate reorder points as a multiple of average demand – typically some version of average usage times a lead-time factor, plus a safety buffer. That’s not a flaw hiding in the configuration. It’s the standard method, and it works fine for items that sell about the same number of units every week.
HVAC inventory doesn’t do that. A condenser fan motor might move four units a week in March and thirty a week in July, then drop to two in November. An average of those weeks isn’t a forecast – it’s a number that’s wrong in both directions at once. Averaging-based reorder logic overestimates the reorder point in the slow months, tying up cash and space on stock that won’t move, and underestimates it heading into peak, which is the stockout you’re trying to avoid in the first place. That’s not a competitor’s criticism of P21. It’s how Epicor itself describes the limitation of multiples-of-average math on its own product page for the tool built to fix it.
What P21 does well on its own
This isn’t an argument that P21’s inventory handling is thin. Multi-location replenishment, transfer orders between branches, and purchasing workflows that treat four or forty stocking locations as the normal case rather than a special configuration – that’s native, and it’s real infrastructure most starter systems don’t have. The gap isn’t visibility. It’s forecasting method. P21 tells you what’s on hand and what’s on order, accurately, in real time, across every branch. It does not, on its own, tell you that the average it’s reordering against stopped being true three weeks ago.
Where the seasonal read actually comes from
Epicor’s answer to that gap is a separate product: Inventory Planning & Optimization (IP&O). It runs alongside P21 rather than inside it, and its job is specifically the thing static averaging can’t do – it selects a forecasting method automatically based on the demand pattern each item actually shows, whether that’s seasonal, trending, intermittent, or cyclical, and it runs the demand and lead-time scenarios probabilistically instead of averaging them into one number.
Practically, that shows up as a policy per item instead of one formula for the whole catalog: the compressor line gets treated like the seasonal item it is, the slow-moving specialty fitting gets treated like the intermittent item it is, and neither one is quietly stealing capacity from the other’s reorder math. The 2026.1 release adds a visual read on the same thing – a sawtooth view of on-hand inventory against replenishment orders over time, for one item or a group, without leaving the workflow to build the chart by hand.
The question worth asking before next peak season
Not “does our ERP handle seasonality” – every vendor page says yes to that question. The one worth asking is narrower: is the reorder point on your fastest-moving seasonal SKUs coming from an average, or from a policy that was built to expect the swing? If it’s the former, the numbers eventually tell on themselves – as a stockout in June or a warehouse full of dead stock in November. Usually both, on different items, in the same quarter.
One ask
If you’re running P21 today and want an honest read on whether your current reorder setup is built for a flat catalog or a seasonal one, that’s a 30-minute conversation, not a sales process.
