Summarize the article:
Kentico Xperience 13 (KX13) reaches End of Technical Support on December 31, 2026. The support date itself does not disable the application, but vendor fixes stop and the risk becomes yours.
Your first real deadline may be earlier. If your live site runs on .NET 8 or .NET 9, those reach end of support on November 10, 2026; current KX13 releases support .NET 10, so this is a runtime check, not a universal cutoff.
The right decision depends on exposure, data sensitivity, customization, and how long you need an interim period — not on one rule for every site.
Moving to Xperience by Kentico (XbyK) is closer to a replatform than an in-place upgrade. The Migration Tool moves data, including Page Builder content; custom code and features are adapted, reimplemented, or rebuilt depending on the feature.
Interim measures buy time. They do not restore vendor support, and some add rework later.
This is a decision guide, not an end-of-life alarm. KX13 won't switch off on January 1, 2027, but "it still runs" is not the same as supported, secure, or compliant. Below are the changes, the costs, what to do if you can't migrate in time, and how to set up your own site.
No. KX13 is a customer-hosted application, and the End of Technical Support date does not switch it off. Continued operation depends on your license model: a perpetual license permits indefinite use, while a subscription license stops after expiration and a grace period.
What ends is Kentico's obligation to keep it running. From January 1, 2027, Kentico ceases "all support, maintenance, updates, releases, hotfixes, patches, repairs (including security repairs), and any other services" for KX13, and states that continued use is "at your sole risk". An implementation partner can still diagnose or adapt custom code; what disappears is the vendor patch pipeline for the product.
That splits four things that used to travel together. A site can be running: the application still executes, regardless of vendor status. It is no longer supported: Kentico stops fixing, patching, or maintaining KX13. It cannot be assumed secure: a newly disclosed vulnerability in KX13 product code in 2027 has no vendor patch. And it is not automatically compliant: whether an unsupported system meets your obligations depends on your industry and your auditor or insurer, not on the software itself. A security review asks whether it's patched; an auditor or insurer asks whether an unsupported system meets your obligations.
Risk depends on your implementation and splits into dated deadlines, slow erosion, and operational exposure. Not every site carries all three at the same intensity — a small marketing site on a stable stack may barely notice the first, while a commerce platform on shared infrastructure feels all three quickly. The point isn't that every KX13 site is in immediate danger; it's knowing which of the three applies to yours.
Dated deadlines vary depending on how the site is built:
| KX13 implementation | Runtime lifecycle to verify |
ASP.NET Core live site | .NET 8 and .NET 9 end support on November 10, 2026 (Microsoft); current KX13 releases target .NET 8–10, so moving to .NET 10 (supported to November 2028) keeps the runtime supported |
MVC 5 live site | Runs on .NET Framework, tied to the Windows Server lifecycle rather than the .NET 8/9 date |
Administration application | ASP.NET Web Forms on .NET Framework, regardless of live-site model; .NET Framework 4.8 is serviced with its Windows host |
Moreover, your database has its own clock. SQL Server 2016 extended support ended July 14, 2026; eligible customers may buy Extended Security Updates for a limited period, so "2016" does not automatically mean "unpatched." SQL Server 2017 follows on October 12, 2027, and SQL Server 2019 runs to January 2030 (Microsoft).
Slow erosion follows the longer you wait: unpatched vulnerabilities can accumulate in product code and dependencies, and engineers fluent in KX13's Web Forms admin and macro engine become harder to source.
Operational exposure is the part teams underweight. Without a vendor to escalate to, your rollback and disaster-recovery plan becomes the safety net for any failed change. Testing also has to be scaled to the change: smoke testing after certificate and infrastructure changes, targeted regression after dependency updates, and broader regression for runtime or platform changes.
"On paper, KX13 keeps running after December 2026. In our assessments, the problem is rarely the CMS itself — it's the accumulated custom code. When a portal has dozens of custom widgets and integrations wired into the CMS, every unpatched dependency becomes something you own, with no vendor fix behind it. We check the live-site runtime and the widget inventory before anyone discusses a timeline."
— Roman Hutnyk, CEO of Bits Orchestra and a former Kentico MVP
No, not in the sense the question implies. Hardening reduces exposure. It does not replace product support, and the two are not equivalent.
Restricting the administration application — VPN, IP allow-list, or a private endpoint — removes a major admin-side attack surface and, for a public KX13 site, it is one of the highest-value controls. A web application firewall, rate limiting, network segmentation, and monitoring reduce known-pattern attacks and shorten detection time.
Where a vulnerability sits on a path that must stay public, such as a member login or a public form, a firewall rule may serve as a compensating control, but it is weaker than a fix. A vulnerability in an unsupported runtime, or in a dependency you can't upgrade without a vendor release, is harder still. Migration is what removes the unsupported-product condition over the long term; it is not the only response to any single vulnerability, but it is the only one that resolves the underlying status.
Cost sits in four places. Direct costs are budget line items: hosting, monitoring and firewall tooling, planned maintenance, and regression testing. Hidden costs surface through delay: the scarcity premium for KX13 specialists, staging upkeep, and incident response without a vendor path. Risk-exposure costs are carried as probability — a breach, insurance friction, business interruption. Opportunity costs are features not shipped during a freeze.
Migration cost varies by system. Bits Orchestra publishes reference ranges of roughly:
The lowest tier reflects a straightforward version upgrade, not a full architectural KX13-to-XbyK migration, which lands higher. Migration also does not end lifecycle management: XbyK uses subscription licensing and a rolling release-support policy.
Treat each option as a temporary, conditional decision with a review date.
Option | What it achieves | What it leaves open |
Enhanced monitoring | Faster detection | The vulnerability itself |
Isolate the administration app | Removes a major admin-side attack surface | Public login and form paths |
Freeze non-essential development | Stops adding surface; stable scope | Existing exposure unchanged |
Move runtime to .NET 10 (before Nov 10, 2026) | Removes the .NET 8/9 deadline | May require the latest KX13 Refresh and hotfix |
Re-host, keep the CMS | Improves infrastructure posture | KX13 support status is untouched |
Temporary maintenance agreement | Assigns incident ownership | No vendor product hotfixes |
Migrate critical sections first | Removes risk on migrated parts | Coexistence of two platforms |
Two of these are commonly oversold, not because the mechanics are wrong, but in how they get pitched internally. Re-hosting is often framed as "addressing the KX13 risk," but it changes nothing about the support clock — it's infrastructure work wearing a risk-management label. A maintenance agreement is called a "safety net," but it catches incidents as they happen, not the absent vendor patch upstream. Both belong inside a bridge, not a permanent answer, and neither substitutes for a documented, time-boxed decision.
Yes. Migration and redesign are separate decisions, and separating them is usually what makes the deadline reachable. XbyK is a rewritten platform on ASP.NET Core with a React-based administration. In a default migration, KX13 sites become website channels and KX13 page types become page content types; structured content moves into a website channel, and remodeling selected content into reusable Content Hub items is an optional, later decision rather than a requirement for the migrated solution to run — with one exception noted below for media assets.
What the Migration Tool carries, and what it doesn't, sets the scope:
The tool requires a source instance on KX13 Refresh 5 (hotfix 13.0.64) or newer, and the supported object set changes between releases — confirm both at project time.
A Bits Orchestra project shows the shape of this. A building-materials manufacturer ran a KX13 dealer portal for 350+ distributors, with 12,000+ SKUs and 80+ custom widgets on a .NET Framework monolith, and needed to move without a greenfield rebuild. The team used the Migration Tool plus a custom script for non-standard content, ran three rehearsal migrations, and cut over in a single weekend with KX13 kept read-only for 48 hours as a fallback. Catalog page loads improved from roughly 3.1s to 2.0s. The redesign came later, as a separate decision.
For a step-by-step view, see how to upgrade from KX13 to Xperience by Kentico.
For most projects, a parallel build — developing XbyK while KX13 stays in production — is the safer default; a production-phased rollout is the riskier exception, not a synonym for "phased." The terms are often confused. A parallel build carries low coexistence risk and suits most projects. Rehearsal migrations repeat the data move in non-production before cutover. A production-phased rollout, where both platforms serve live traffic for a period, is different: it requires traffic routing, shared authentication, content ownership rules, and synchronized data, and it can increase total effort rather than lower it. For a smaller, self-contained site, a single cutover is often the lower-risk choice.
Rate your site on exposure, data sensitivity, customization, current runtime, and how long an interim period must last.
Path | When it fits |
Continue temporarily, controlled | Low exposure, mostly static, admin-restricted, non-sensitive data, current release and runtime. Documented as a time-boxed, accepted risk. |
Stabilize, then phase the migration | Business-critical with real exposure; isolate the admin, monitor, freeze, move to .NET 10 if needed, migrate high-risk sections first. |
Migrate now | Public authentication, forms, uploads, or payments with sensitive data; or on .NET 8/9 with no path to .NET 10; or under compliance or insurance pressure. |
Replace or redesign | The content model, customizations, or integrations no longer fit the business, so a like-for-like move would carry the constraints forward. |
Every path starts from the same input — a dependency audit. Before requesting an estimate, prepare:
That inventory is what separates a defensible hold from an unmanaged one, and a realistic estimate from a guess— it's also the starting point for choosing a Kentico development partner, since a vendor's ability to scope from this list, rather than a generic quote, is the fastest signal of competence. Bits Orchestra's Kentico migration services start with that dependency assessment before any implementation scope is set.
No. KX13 is customer-hosted and keeps running, subject to your license model. What ends is vendor support — no more hotfixes, patches, or security repairs — and continued use is at your own risk.
It can be hardened, not made equivalent to a supported product. Restricting the admin, a firewall, and monitoring reduce exposure, but a vulnerability in product code on a public path or an unsupported runtime has no vendor fix.
Yes. The Migration Tool moves structured content and Page Builder data, so you can move to a like-for-like XbyK build first and redesign later. Custom code, commerce, and integrations are adapted or rebuilt.
It depends on customization and integrations. Smaller, lightly customized sites can move in about 4–8 weeks; large or enterprise projects with widgets, commerce, and integrations to rebuild often run 6–12 months across phases.