What Epicor Actually Announced
Epicor confirmed in early January that it is ending on-premises development for its three flagship ERP lines: Kinetic, aimed at mid-market and upper mid-market discrete manufacturers with complex shop-floor needs; Prophet 21, used by wholesale and industrial distributors managing inventory, pricing, and order fulfillment; and BisTrack, built for building materials, lumber, and construction supply distributors. Future innovation, including every new AI feature, moves exclusively to Epicor Cloud going forward. The company frames this as the arrival of what it calls the first ERP AI agent built with outcomes-based pricing, a capability it is positioning as available only to customers who have already made the move to its cloud platform.
This is not a quiet deprecation notice buried in a release notes document somewhere. Epicor set specific, product-by-product final release dates and support windows running from 2026 through 2029, giving customers a concrete deadline rather than an open-ended recommendation to modernize whenever budget allows. More than 20,000 businesses already run on Epicor Cloud, a figure the company is using as evidence that the migration path is proven at real scale across its customer base, not an experimental option it is pushing customers toward without a track record behind it.
The Clock, Product by Product
The specific dates matter enormously for budget planning, since they define exactly how much runway each product line still has left. Kinetic's final on-premises release lands around January 2028, with active support running through December 31, 2029, and sustaining support starting January 1, 2030. Prophet 21's final release follows around May 2028, with the same active-to-sustaining support pattern ending June 2029. BisTrack's web browser and API components mirror that same 2028 to 2029 window, giving most customers a multi-year, but still finite, runway to plan and execute a full migration.
BisTrack Desktop customers face the tightest runway of the group: a final on-premises release around December 2026, with active support ending December 31, 2028, and sustaining support picking up from January 2029. An older BisTrack UK version dating back to 2017 loses active support entirely by the end of this calendar year, with only sustaining support following from January 2027. For CIOs running any of these specific product versions, the gap between now and the first real support cliff is measured in months, not years, which should already be shaping 2026 and 2027 capital budget conversations rather than waiting for a future planning cycle.
Why Architecture, Not Policy, Is the Real Barrier
Manish Jain, principal research director at Info-Tech Research Group, frames this shift bluntly: vendors are taking the on-premises alternative off the table entirely rather than simply making cloud more attractive on price. He argues this is a long-term operating model change disguised as a hosting decision, and that regulated industries are not blocked from cloud ERP by policy. They are blocked by architecture, meaning compliance becomes an engineering problem to solve rather than a reason to delay indefinitely.
Robert Kramer, VP and principal analyst at Moor Insights and Strategy, adds a practical nuance: Epicor is not turning off existing on-premises systems overnight, and customers keep receiving support through the published windows. The real change is that staying on-premises becomes what Kramer calls a supportable maintenance decision rather than a growth one. Every new capability, every AI feature, and every competitive advantage Epicor ships from here forward lands in the cloud product first and exclusively.
The Vendor-Dependency Blind Spot
Jain's sharpest warning concerns resourcing rather than technology itself. He argues many CIOs have not fully resourced the shift to managing vendor dependency as its own discipline, treating this announcement as a line item in an IT project plan rather than a multi-year change management effort that touches procurement, finance, field operations, and often the shop floor directly. That gap between the technical migration plan and genuine organizational readiness is consistently where ERP transitions run over budget, slip their schedule, and lose executive sponsorship partway through.
This gap is especially acute for the manufacturers and distributors Epicor serves, many of which are mid-market, private-equity-backed, and running lean IT teams relative to their revenue and operational complexity. A forced multi-year cloud migration now competes directly with other capital priorities on a finite budget, and the businesses with the least bandwidth to absorb a disruptive ERP transition are frequently the same ones least prepared for the governance and change management work this kind of shift actually demands of them.
Who Feels This First
Highly regulated and data-sensitive organizations face the steepest climb, since they may need hybrid, private, or sovereign cloud configurations that take meaningfully longer to architect and validate than a standard public cloud migration would. Heavily customized instances carry a similar risk profile: years of bolt-on code and workflow changes rarely transfer cleanly to a standardized cloud platform, and discovering the scale of that gap mid-migration is far more expensive, and far more disruptive to the business, than mapping it honestly before the project ever gets a budget line and a start date.
Distributors on BisTrack Desktop sit closest to the deadline and should already be furthest along in planning, given the December 2026 final release target bearing down on them. Discrete manufacturers on Kinetic and industrial distributors on Prophet 21 have more apparent runway through 2028, but that runway shrinks fast once a migration project actually gets underway, particularly for organizations that have never run a full ERP replatforming before and will need to budget for external systems integrator support to execute one safely, on schedule, and without disrupting daily operations.
What to Do With Your Budget Cycle Now
The practical move for any CIO on these platforms is to treat this announcement as a forcing function for a conversation that was probably already overdue. Map your specific product version against Epicor's published support windows, identify whether your instance carries heavy customization or regulatory constraints, and get a realistic cost and timeline estimate from a systems integrator before the next budget cycle locks in priorities for reasons unrelated to this deadline.
The organizational lesson extends well beyond Epicor and this specific product family. Vendors across the broader ERP market are increasingly using hard support sunsets rather than pricing incentives to push customers toward cloud platforms, which means the next system you depend on could issue a similarly abrupt notice with relatively little advance warning. Building vendor-dependency review into your regular governance cadence, with a standing owner and a standing budget line, rather than reacting deadline by deadline as each notice lands, is the only durable way to stop announcements like this one from always arriving as a surprise to your roadmap.
