When the Vendor Disappears
A building that rents its behavior from a subscription inherits a dependency it cannot see — until the invoice stops, the server shuts down, or the company pivots.
In 2023, a widely used visitor management service shut down with four months' notice. Buildings that had integrated the service into their lobby operations — printing badges, logging arrivals, notifying residents — had four months to find an alternative, migrate their data, retrain their staff, and reconfigure their hardware. Some of the buildings had been using the service for six years. Their visitor logs, configuration history, and operational workflows were hosted entirely on servers they did not control.
This was not a hack. Not a breach. Not a failure. The company decided to close the product. The buildings discovered that the infrastructure they depended on was not infrastructure at all — it was a subscription. And subscriptions end.
The spectrum of dependency
Not every vendor shutdown is dramatic. Most dependencies fail gradually. A company raises prices. A company is acquired and the new owner deprecates the product. A company pivots to a different market and stops investing in the residential feature set. A company's engineering team shrinks and response times grow from hours to weeks. A critical integration breaks and the vendor's support ticket sits open for three months.
These are not catastrophes. They are erosions. The building's operations degrade slowly enough that no single event triggers a response — but the cumulative effect is a system that no longer behaves as it did when it was selected.
The data question
When a vendor-hosted service fails or shuts down, the building faces a question it should have asked on day one: who owns the data?
Resident records. Payment histories. Maintenance logs. Delivery records. Access audit trails. Visitor logs. In most vendor arrangements, this data is stored on the vendor's servers, in the vendor's format, accessible through the vendor's interface. When the vendor disappears, the data either disappears with it or is returned in a format that no other system can read without significant effort.
The ADAG Lifecycle × Reliability expectation is explicit: when a building system is changed, replaced, or upgraded, the resident's account data, payment history, and documents are preserved without gap or loss. This expectation is easy to meet when the building controls its data. It is difficult when the data lives in a system the building rents.
The evidence is not theoretical
Between 2020 and 2025, the proptech industry experienced a wave of consolidation, shutdowns, and pivots that left buildings stranded. Access control companies acquired by competitors who discontinued the product line. Visitor management services that shut down entirely. Delivery management systems that raised prices beyond what residential operators could justify. Amenity booking services that pivoted to commercial office and stopped supporting residential features.
Each shutdown followed the same pattern: the building had no local copy of its operational data, no contractual guarantee of data portability, and no technical ability to migrate to an alternative without a gap in service. The resident experienced the gap as: the gym booking system changed and lost all my reservations. The delivery notifications stopped for two weeks. The maintenance history for my unit disappeared.
What the standard requires
The standard does not prohibit vendor-hosted services. It requires that the building's behavior is not contingent on any single vendor's continued operation. This means: operational data is retained locally or in a format the building controls. System transitions preserve resident records. No lifecycle event — payment history, credential activation, household composition — is stored exclusively in a system the building does not own.
A building that meets these requirements can survive a vendor shutdown without the resident noticing. A building that does not has delegated its reliability to a company whose incentives may not include the building's continued operation.
The question is not whether a vendor will disappear. It is whether the building has been designed for the day it happens.