Four streams pass through audit points between a disconnected platform and a new stable platform, where they come together.
Platform Rulesthrough Where the Work Breaks

A Small Service for Auditing App Migrations After a Platform Shutdown

As Gadget winds down, there is room for a focused migration service that finds and moves the points where sign-in, data, notifications, and customer guidance can break—instead of rebuilding an entire app.

Published 2026. 11. 17.

What closes before the service does

Gadget announced plans to end its service on November 9, 2026. Existing apps can run until March 31, 2027, but the creation of new apps stopped with the announcement.

Apps that have not yet released a customer-facing version can make their first production deployment only until the end of November 2026. A production deployment is the version that customers actually use.

Customer support will be available only until the end of 2026. Because the channel for questions and help will effectively narrow three months before the full shutdown, app operators are safer deciding what must move and who owns it before year-end.

More than screens move through Gadget. Customer accounts and permissions, saved data, scheduled jobs, automatic signals from external services, and rules for rerunning failed work all run through it too.

The company said it would provide migration guidance and tools. But rather than wait for those tools, operators should first write down which data they need to extract, where to extract it from, and what they will compare it with in the new environment.

Move the connections between work, not just the app

Imagine a five-person team operating an order-notification app for Shopify merchants. When an order appears in a customer store, Gadget saves the order details, selects notification copy, sends it to a messaging service, and saves the result again.

The team views store settings in the Gadget admin screen and checks connection status in the Shopify admin screen. When a customer says a message was missed, the team switches between the messaging-service screen and the customer-support tool, manually copying the store ID, order ID, and send time.

Once migration starts, movement between tools increases further. The team downloads existing data into files, loads it into a new data store, makes a list matching old store IDs to new ones, and records missing items in another file.

Sign-in is separate work. Email addresses and permissions may be transferable, but copying an existing browser sign-in session as-is is not safe. After matching administrators and regular users, the team needs to use real accounts to confirm that signing in again or resetting a password works properly.

Automated notifications do not end with copying message settings. The migration also needs to include future scheduled sends, items already waiting for a retry after failure, customer consent to receive messages, and the marker that prevents the same notification from being sent twice.

In the new environment, the team can copy data first and block real sends while comparing results. It can feed the same order into both the old and new environments, compare saved data and notification recipients, then stop the old schedules and turn on the new ones to avoid duplicate sends.

Some work still remains with people. They must decide permissions for exception customers, correct badly organized data, and explain features that customers need to reconnect. The value of a migration service is not making those decisions for them. It is collecting every item they need to decide in one place so none is missed.

Others moved the connections first

WhatFlow, based in Pakistan, operates WhatsApp order notifications and cart-recovery services for Shopify merchants. It charges customers a monthly subscription and fees for message overages. As Gadget costs and processing-volume issues grew, it moved to several operational tools from Cloudflare.

WhatFlow moved its official WhatsApp connection first and preloaded data into the new environment before final verification and cutover. The company says it moved more than 22,000 connected sessions without requiring merchants to reinstall the app or reconnect WhatsApp, but this is a company-reported figure and needs independent verification.

DataStax Academy in the United States moved to FusionAuth after Stormpath, the sign-in service it used for education and product downloads, was shut down. FusionAuth provides companies with sign-in management software, an operating environment, and technical support for a fee.

The migration included not only the user list but also password-verification methods, roles, download permissions, and authentication information between apps. FusionAuth's case material says the live authentication cutover took less than 10 minutes, but this is a vendor case study, so performance and savings claims need independent verification.

Import2 in the United States moves historical data from customer-support tools such as Help Scout and Zendesk for customers, charging a one-time fee based on the amount of data and the tools being connected. Help Scout connected Import2 because it had only two migration paths that it supported directly.

As a result, the number of source tools from which Help Scout customers could migrate grew to more than 25. Its claim that migrated customer numbers increased 2.3 times is a company-reported figure, but it shows that instead of building a new product, a team can turn one task—making old conversations appear unchanged in a new interface—into a service.

Things you can build now

1. Migration item map

  • What it does: Takes a Gadget app's screens, data, sign-in, scheduled notifications, and external connections, then puts missing migration items and owners on one page.
  • Who uses it: A team of three or four people running a Shopify app on Gadget without a separate technical owner.
  • Why now: Support ends before the whole service does, so teams need to document their connection structure while they can still ask questions.
  • First screen: Four sections for sign-in, stored data, automated notifications, and external connections, with export status, owner, and test date for each item.

2. Sign-in migration practice space

  • What it does: Loads users and permission lists into a new sign-in environment for testing, then compares results for administrators, regular users, and deactivated accounts.
  • Who uses it: An app operator with paid members or store administrators, where incorrectly migrated permissions could expose customer data.
  • Why now: Existing sign-in sessions are difficult to copy directly, and preparing password-reset guidance can reduce questions on the actual cutover day.
  • First screen: The existing sign-in method, user types, and three test accounts, alongside results for signing in again, changing a password, and checking permissions.

3. Duplicate notification checker

  • What it does: Sends the same order signal to the old and new environments while blocking actual customer sends, then finds differences in recipients, copy, and scheduled times.
  • Who uses it: An operator of a Shopify app that automatically sends order confirmations, delivery updates, or cart notifications by SMS or WhatsApp.
  • Why now: Copying scheduled jobs wholesale can cause not only missing messages but also duplicate sends to the same customer.
  • First screen: For each order received today, show the old result, new result, scheduled send time, and missing or duplicate warnings side by side.

4. Customer-by-customer cutover guide room

  • What it does: Separates customers by whether they need to sign in again, reconnect a store, or check settings, and manages guidance delivery and completion status.
  • Who uses it: An order-automation app team where notification copy and connected accounts differ by customer store.
  • Why now: If every customer receives the same shutdown notice, the actions they actually need to take get buried and questions pile up on cutover day.
  • First screen: Next to each customer name, show four statuses—automatic migration, sign-in required, manual connection required, and confirmed complete—along with the guidance to send.

Why this matters where you are

A platform shutdown can break sign-in, stored data, automated work, and customer communication before an app stops running. Check whether the platforms your product depends on let you export those parts, test them in a new environment, and identify which customers must take action. The exact tools and support timelines will differ by market and platform, but mapping those connections can reveal a narrow migration task worth solving.

What to check in 30 minutes today

Choose one real order or customer action and follow it from the moment it happens to the completed notification. Write down, in order, every tool you open and every value you copy by hand. If you switch tools three or more times, or if the owner and test method for sign-in, data, or notifications is missing, it may be worth testing a small migration service that solves only that break point.

Sources

8 sources

Every fact in this article came from the pages below. Check them yourself.

A Small Service for Auditing App Migrations After a Platform Shutdown | Prometheon