If you own ArcGIS applications, you have probably been told that Web AppBuilder is going away. That is true, but almost every summary of it makes one of two mistakes, and both of them lead to bad planning decisions.
The first mistake is treating "you can no longer edit it" and "it stops working" as the same event. They are more than six months apart, and the earlier one is the one that constrains you.
The second mistake is bigger: applying the ArcGIS Online schedule to ArcGIS Enterprise. Enterprise has no announced shutdown date at all. If you run your own portal and someone has told you your applications die in 2027, they have not checked.
We are an Esri partner and we do this work, so treat this as interested rather than neutral. But the dates below come from Esri's own documentation, linked at the bottom, and you should check them yourself rather than take our word for it.
Web AppBuilder in ArcGIS Online
Three milestones, all published by Esri:
Q1 2026: no new applications. Already passed. You cannot create a new Web AppBuilder app in ArcGIS Online.
Q4 2026: editing ends. You will no longer be able to update existing Web AppBuilder applications. The apps keep running, but they freeze in whatever state they are in.
Q2 2027: applications stop functioning. Existing Web AppBuilder apps are retired and no longer work.
Note that this schedule was extended. An earlier plan had retirement landing in Q1 2026, so if you looked at this a year ago and filed it as urgent, you may have since heard it was pushed back and stopped worrying. The extension is real, but the Q4 2026 editing freeze is now roughly a quarter away.
Why the editing freeze matters more than the shutdown
The shutdown date is the one that gets the attention, because it is the one where something visibly breaks. But the editing freeze is the constraint that actually shapes your project.
After the editing freeze, any change to a frozen application has to be made somewhere else. A new layer, a corrected popup, a rebranded header, a fix to a broken service reference: all of it now requires the replacement to already exist. So the freeze does not just stop you editing. It converts every subsequent change request into a migration request.
That is the practical deadline. If an application will need any change between Q4 2026 and Q2 2027, its real deadline is the earlier date.
Web AppBuilder in ArcGIS Enterprise is a different situation
This is where most of the confusion sits, and it matters because the two platforms are on completely different tracks.
Web AppBuilder in ArcGIS Enterprise has been deprecated since March 2023, and ArcGIS Enterprise 11.5 is the final release that includes it. But there is no announced date on which Enterprise-hosted Web AppBuilder applications stop working.
What that means in practice: your constraint is your own upgrade path, not a countdown. Your applications keep running as long as you stay on a version that ships Web AppBuilder. At some point you will want or need to upgrade past 11.5, and that upgrade is the moment Web AppBuilder leaves. You control the timing, which makes this a planning question rather than an urgent one.
It is worth planning properly. It is not worth panicking about, and anybody telling you otherwise is either confused or selling.
Developer Edition is already gone
Web AppBuilder Developer Edition was retired in July 2024, alongside the retirement of ArcGIS API for JavaScript 3.x, which Web AppBuilder is built on.
If you have custom widgets, this is the part that deserves your attention more than any of the dates above. Custom widgets were written against a JavaScript API version that is no longer supported, using a development tool that no longer exists. They do not migrate by reconfiguration. Whether the source code still exists, and whether anyone can still explain what it does, tends to determine the shape of an entire migration project.
That question is worth answering now, regardless of which platform you are on, because it is the one with the longest lead time.
A quick summary
| Product | Status | What happens |
|---|---|---|
| Web AppBuilder in ArcGIS Online | Retiring on a schedule | No new apps since Q1 2026. Editing ends Q4 2026. Apps stop working Q2 2027. |
| Web AppBuilder in ArcGIS Enterprise | Deprecated | Deprecated March 2023. Enterprise 11.5 is the last release including it. No shutdown date. |
| Web AppBuilder Developer Edition | Retired | Retired July 2024, with ArcGIS API for JavaScript 3.x. |
Not everything should become an Experience Builder app
Experience Builder is the natural destination and often the right one. Treating it as the only answer is how migrations get expensive.
Across the public Web AppBuilder applications we have surveyed, the majority are read-only viewers: a map, a few layers, a legend, a search box, an attribute table. Rebuilding that as a full Experience Builder application is usually more app than the job needs, and more to maintain afterwards.
Worth considering per application:
- ArcGIS Instant Apps for read-only viewers with a small widget surface. Much less to maintain.
- ArcGIS Dashboards where the real job is monitoring and summarising rather than map interaction.
- ArcGIS StoryMaps where the application is actually communicating a narrative.
- Experience Builder where there is genuine workflow depth, editing, or many widgets.
- A custom application where custom widget behaviour has no configurable equivalent and genuinely matters.
- Consolidation. Organizations frequently have several near-identical viewers that differ only in extent or layer selection. Merging is usually cheaper than rebuilding each one.
- Retirement. Some applications are simply no longer used. Finding those is often the fastest saving available.
Where to start
Three questions, in this order, and none of them need a consultant:
- Which platform are you on? ArcGIS Online means you have a dated deadline. Enterprise means you have an upgrade decision. This changes everything downstream.
- How many Web AppBuilder applications do you actually have? Most organizations underestimate this, because applications accumulate quietly. Public ones are easy to count. The internal ones behind your login are usually the majority, and usually the complicated ones.
- Do any of them use custom widgets, and does the source code still exist? This is the question with the longest lead time and the one most likely to change your plan.
If you want a starting point, we built a free migration readiness assessment: six questions about your portfolio, and it tells you which of these three is actually your constraint. It needs no email address to show you the result, and if you are on Enterprise it will say so rather than selling you a deadline you do not have.
We also published what we found inside 14,978 real Web AppBuilder applications, which is useful for calibrating how big a job yours is likely to be.
If the answer is that you want this done properly across a portfolio, that is what our five-day modernization Blueprint is for.
Sources
Check these rather than trusting this summary, and check them again before making a decision, because Esri has revised this schedule once already:
- ArcGIS Web AppBuilder retirement, Esri documentation
- ArcGIS Web AppBuilder in ArcGIS Online Retirement Extended, Esri blog
- ArcGIS Web AppBuilder life cycle, Esri Support
Mapular is an Esri partner with the Nonprofit Specialty. We built the UNESCO World Heritage Site Navigator with Esri Deutschland on ArcGIS Experience Builder, Dashboards and Notebooks, monitoring more than 1,000 inscribed properties.
