What Epicor's 2028 On-Premise Deadline Actually Means for Your Next Software Decision

More than one Epicor customer has started calling the 2028 on-premise deadline a rug pull. That word keeps showing up in shop-floor forums and LinkedIn comments, and it's not really an exaggeration. If you built your operation on-prem, on your own server, on your own timeline, you didn't sign up for a date that isn't yours. Most people reading this are still running Kinetic today, so we're not here to pile on Epicor or a system that's probably still doing its job fine. We're here because if you own the P&L, you've likely already had the quiet version of this conversation with your controller or your IT person: what do we actually have to do about this, and by when.

What's actually forced by 2028

Here's the part that isn't optional. Epicor has told on-premise Kinetic customers that standard support ends in 2028. After that date, the patches stop: no more security updates, no more tax table changes, no more compliance fixes for the version you're running. Staying on an unsupported on-prem build past that point isn't really a decision, it's a risk you'd be choosing to carry, and most auditors, insurers, and customers with their own compliance requirements won't let that risk sit quietly in the corner. So the forced move is real: at some point before the deadline, your Kinetic instance moves to Epicor's cloud.

What's still your call

What's not forced is everything past that. Moving your existing Kinetic system to the cloud is one project. Ripping out Kinetic and shopping for a different ERP entirely, whether that's NetSuite, Infor, or something built specifically for job shops, is a separate project with its own cost, its own risk, and its own timeline. Too many owners hear "2028 deadline" and file it under "we need new software," when the actual requirement is closer to "we need new hosting." Those aren't the same decision, and you don't have to make both of them on the same clock.

The real cost question

This is usually where the epicor kinetic cost conversation gets confused. The migration itself, moving a well-configured Kinetic environment from your server room to Epicor's cloud, is typically the smaller number on the page. The bigger number, the one that actually varies shop to shop, comes from what you choose to do around the migration: how many customizations you're carrying, how much data cleanup is overdue, how much retraining your team needs, and whether you use the moment to fix workflows nobody's touched since the system went in. Get a real quote before assuming the worst. A lot of the sticker shock people report on comes from bundling a full reimplementation into what could have been a straight lift and shift.

The decision actually in front of you

Strip away the deadline pressure and there are really three paths. Eat the migration cost now, move to Kinetic in the cloud with as little disruption as possible, and keep running the shop the way you already run it. Use the forced pause as a real chance to evaluate alternatives, but only if there's an actual reason Kinetic isn't fitting your shop today, not because a deadline made you nervous. Or take the option most owners skip: use this required project to fix what's actually broken, which usually isn't the ERP software itself. It's how hard it is to get a straight answer out of it.

Don’t just move the same seven screens

Ask anyone running Kinetic day to day where the real friction lives, and it's rarely the software crashing. It's the five minutes spent clicking through job costing, routings, and PO history to answer a question a customer just asked on the phone. Migrating to the cloud and calling it done just moves that same friction onto a different server. If you're already touching the system for a required move, it's worth asking whether you also want the thing that actually changes how your team works day to day: straight answers pulled from the ERP data you already have, in plain English, without anyone hunting across screens to find them.

The 2028 deadline is real, and it forces exactly one decision: get off unsupported on-prem before the patches stop. Everything else, replacing the system entirely or fixing how your team actually gets answers out of it, is yours to decide on your own schedule, not Epicor's.