Centralized Access Control: How It Works and When It Fits

Profile image for Nick Manzolillo
Nick Manzolillo
Updated 17 min read
Used by more than 1 million, including the most trusted names in real estate
related-logo-png lincon-property-logo-png lennar-logo-png CA-ventures-logo bozzuto-logo-png

Key takeaways:

  • Centralized access control brings staff administration into one platform, while controllers may still make entry decisions locally.
  • A working credential during an outage does not prove that a new permission or access removal has reached the door.
  • Choose centralized, local or hybrid administration based on documented door behavior, comparable costs and a representative pilot.

 

 

Centralized access control lets you manage credentials, permissions and entry records for multiple doors, gates or properties through one platform. For a regional property manager, multifamily centralization can mean handling resident, staff and vendor access without repeating changes in several separate systems.

But one dashboard does not tell you what happens at a disconnected building. Existing credentials may keep working even when a new permission cannot reach the door. Before replacing your current systems, check both promises: continued entry and delivery of permission changes.

Understanding that difference helps you decide where shared administration makes sense, where local control should remain and what a pilot needs to prove.

This guide will answer:

 

Watch how ButterflyMX works:

 

 

What is centralized access control?

Centralized physical access control uses one platform to manage credentials, permissions and entry records across multiple entry points, which may span several properties. Decentralized systems require separate administration, as explained in Avigilon’s comparison of centralized and decentralized access control.

A single community can be centralized if staff manage its entrance, garage and amenities through one platform. By contrast, electronic credentials alone do not make a system centralized. If you must update a resident’s permissions in several separate systems, administration remains divided.

This guide concerns doors, gates and other physical entry points, not IT account and application access management. Cloud-managed versus on-premise is another distinction: it describes where software is hosted, not whether staff share one administrative platform. As ButterflyMX’s guide to scalable access control explains, cloud hosting is not mandatory for scalability.

Your decision is therefore more specific than whether to buy cloud software. You need to know whether shared oversight is worth the transition and whether the proposed system can enforce your access rules under the conditions your properties face.

 

How permissions reach a door

The platform where you manage permissions is not necessarily where the system decides to unlock a door. To understand that difference, start with the roles of the credential, reader, controller and administrative platform.

 

From credential to entry record

A typical physical access request follows this sequence, described in Avigilon’s system overview and Kastle’s apartment access control guide:

  1. A person presents a supported credential, such as a fob, card or mobile credential.
  2. A reader receives the credential information and passes the request onward.
  3. A controller or platform checks the applicable access rules.
  4. The system grants or denies entry and operates the lock when access is allowed.
  5. The system records an event, such as the credential, entry point, time and outcome.

The reader receives the request. The controller or platform checks it against the rules. Staff use the administrative platform to manage those rules. A product may combine functions, but these jobs remain distinct.

Central management can coexist with distributed access decisions, meaning decisions made locally at controllers. For example, Verkada documents local use of previously received rules on its AC41 controller. Stored events upload after reconnection. Central administration therefore does not mean every entry needs a live cloud check or every event appears in the shared platform immediately.

 

Roles, time windows and location scope

Permissions define more than whether a credential exists. SmartRent describes roles, time windows and location scope as conditions used in property access decisions. Together, they define which users may enter, when they may enter and which areas they may use.

Consider a hypothetical policy. A resident receives access to an assigned building and its amenities. A technician receives access to designated properties. A vendor receives access to one service entrance during an approved visit. Under those rules, the resident’s credential would not open another building, and the vendor’s credential would not allow entry outside the approved window. Available location limits and expiration options depend on the product and configuration.

Role-based access control, or RBAC, groups permissions by role and assigns users to those roles. For example, a technician role could carry an approved set of permissions rather than requiring staff to select each permission for every technician. The archived, IT-oriented NIST RBAC project explains this permission model; it does not endorse a physical access product.

When you change a role or time window, two tasks remain separate: you save the change in the platform, and the affected door must receive or otherwise enforce it. That second task becomes especially important during an outage.

 

What continues during an outage?

Losing internet does not necessarily stop entry. However, a property internet outage, local network failure, platform outage and power loss are different conditions. Evidence that a controller handles one does not show how it handles the others. The documented designs from Verkada and Kisi show why the details matter.

 

Existing access is not the same as a new permission

In Verkada’s documented AC41 example, the controller keeps the last configuration it received from the cloud when internet connectivity is lost. An existing employee’s badge continues to work. A badge created for a new employee during the outage does not work until the controller reconnects.

The difference is not simply whether staff created a valid badge in the platform. The disconnected controller has not received the new badge’s permission. When connectivity returns, the AC41 retrieves updates and uploads stored events. This example applies to the documented controller; it is not a rule for every centralized system.

Verkada also states that the last configuration survives a reboot. That means the rules remain stored when the controller restarts, not that the controller operates without power. The example also does not establish when access removal, or revocation, takes effect offline. A newly created badge is not a test of a revoked badge.

 

Offline coverage depends on the design

Kisi documents a different approach: access rights are cached, or stored locally, based on usage patterns. Its documentation describes coverage up to 36 hours and daily synchronization at midnight. Recently added or rarely used credentials may be absent from that cache.

Duration and user coverage must therefore be checked together. A stated offline duration does not help a resident whose credential is not included. These are Kisi-specific details, not a category-wide time limit or a limitation of every offline access control system.

Kisi also describes different connection needs for different access paths. Its phone fallback requires phone connectivity, while local reader-controller communication has local-network requirements. Neither statement establishes operation during power loss.

These two documented examples show why an offline label alone is not enough. The comparison below explains different designs, rather than ranking products.

Swipe to compare →

Documented example What continues and who is covered Duration and synchronization New permissions and revocation Dependencies to verify
Verkada AC41 The controller uses its last received configuration. The documented existing badge works. Configuration survives reboot. Updates are retrieved and stored events upload after reconnection. This example does not establish a general offline-entry duration. The documented new badge waits for reconnection. Offline revocation timing is not established. Confirm the controller, credential and supporting equipment. Stored rules do not establish operation without power.
Kisi documented offline support Usage-based cached rights support entry. New or rarely used credentials may be absent. Up to 36 hours of coverage, with daily midnight synchronization and the stated limits on user coverage. A new credential may not be cached. Obtain revocation timing and confirmation for the proposed configuration. Phone connectivity is needed for phone fallback. Reader-controller communication has local-network requirements. Confirm power-loss behavior separately.

The important comparison is not just whether a door opens offline. Continued entry, eligible users, delivery of permission changes and recovery are separate questions. A working credential does not prove that a new permission or revocation reached that door.

 

Evidence to request before accepting an offline claim

Ask the vendor and installer for written answers tied to the proposed controller, credential and configuration:

  • Which credentials and entry points remain usable under each relevant failure condition?
  • How long does coverage last, and which users might be missing from a cache?
  • How do newly issued permissions reach the affected doors during an outage?
  • When does revocation take effect, and what confirms enforcement at each affected entry point?
  • Which power, internet, local-network and phone connections must remain available?
  • How are stored events recovered, and how is recovery confirmed?
  • Who handles an unresolved access exception, and when is it escalated?

The documentation discussed here does not establish offline revocation timing for either design. That gap is not proof of product failure. It is a reason to obtain an answer for your proposed configuration before accepting it. Decide how quickly you need a permission change to take effect and how you will verify it, rather than relying on a general offline promise.

 

Choose centralized, local or hybrid administration

With those door-level requirements defined, you can compare administration models without assuming replacement is the answer. Avigilon identifies uses for both centralized and decentralized administration, including shared multi-site oversight on one side and independent local operations on the other.

 

Centralized administration for shared oversight

One platform can suit frequent cross-property changes, resident turnover, temporary users and shared records. Staff have a common place to manage policies instead of maintaining separate management profiles.

That may fit a technician who works across several communities or a regional team that reviews access records. But a shared platform does not by itself prove lower costs, fewer incidents or faster verified revocation during every outage. Evaluate those results separately.

 

Separate local administration for independent needs

A remote property with limited connectivity or a genuine need for independent authority may justify keeping local administration. Retain the current system when its documented behavior meets the site’s needs and the burden of migration outweighs the supported benefit of shared oversight.

For example, if a site rarely needs cross-property changes and must keep independent access authority, consolidation may solve little of its actual problem. Still, local administration does not remove power, network or recovery needs. Document those requirements for the retained system, too.

 

Hybrid administration with accountable exceptions

In a hypothetical hybrid arrangement, a portfolio could share administration across most communities while keeping a separately managed site for a documented operating reason. Give that exception an owner, clear approval rules and a review date so it does not become an untracked gap.

This management arrangement differs from local decision-making at a centrally managed door. A controller that applies centrally assigned rules offline can still be part of a centralized administrative system.

Use these questions to choose an initial scope, without assigning a score or an automatic winner:

  • How often do permissions change across properties, and who needs shared oversight?
  • How quickly must new permissions and revocations take effect and be verified?
  • Does documented offline coverage include the relevant credentials and entry points?
  • Where must independent local authority remain?
  • What training, hardware changes and credential transition would migration require?
  • How do itemized costs compare over the same planning period?

Weak internet alone does not decide the result. The proposed configuration may meet your requirements, or it may leave a gap you cannot accept. Shared software also does not require a central staffing team. iApartments describes single-site, regional-pod and portfolio scopes as planning options. Choose responsibilities around the work your team needs to perform, not a required staffing model.

 

Follow access through the resident and vendor lifecycle

Once you have an initial scope, follow the daily work it must support. This hypothetical workflow draws on Kastle’s discussion of move-ins, lost credentials, move-outs and access records:

  1. At move-in, an authorized staff member approves building and amenity access, sets up a supported credential and verifies that the required entry points recognize it. Where separate systems previously required separate updates, shared administration gives staff one place to manage the change.
  2. For a vendor visit, a responsible staff member approves the entrance and time window, checks the supported expiration options and keeps the access record. Automatic expiration and location limits depend on the configuration.
  3. For a lost credential, staff departure or move-out, the authorized person approves access removal. Staff then remove permission in the platform and obtain evidence that the affected entry points enforce the change. Any disconnected-door exception follows the agreed escalation process.

The final step needs particular care. An audit trail showing that a manager removed permission records an administrative action. By itself, it does not confirm that a disconnected door received or enforced that change. Keep those two forms of evidence separate when closing an access-removal request.

A property management system, or PMS, integration can trigger parts of this workflow, but centralization alone does not provide that automation. Kastle discusses lease-linked access changes; ButterflyMX also describes portfolio management and Yardi, Entrata and RealPage integrations in its guide to managing access across multiple properties remotely. Before relying on a lease-triggered action, confirm the applicable integration, trigger and customer configuration.

Even when software starts a change, keep approval and exception-handling responsibilities clear. The workflow still depends on trustworthy administrator access and sound practices at the door.

 

What security risks remain?

A central account or platform can control access across a wider area, so a compromise may affect more properties or permissions. That risk comes alongside the convenience of remote oversight and shared policies. Avigilon and SmartRent discuss cybersecurity and operating tradeoffs, but their guidance does not establish that either model is always more secure.

Ask each provider to document how administrator accounts are protected, how staff privileges are limited, how changes are recorded and how departing staff lose access. Also identify who handles account recovery, service recovery and vendor escalation. Treat these as requirements to verify, not features to assume.

Credential security is a separate decision. The ASIS access control report describes weaknesses in older credential technologies. Ask for an assessment of the credentials proposed for your project. Managing an existing credential centrally does not strengthen the credential itself.

The same report documents tailgating and propped-door problems and emphasizes policies and training alongside technology. Its findings come from a self-selected September 2023 survey of security professionals across industries, sponsored by FacilityOS. The survey does not compare centralized and local systems or provide a current multifamily benchmark.

Your proposal should therefore identify both the people responsible for safeguards and the equipment needed to support them. A software demonstration alone cannot answer those property-specific questions.

 

Assess existing hardware before planning migration

Before choosing replacement equipment, list your controlled entry points, current systems, credential types, responsible vendors and operating problems. iApartments’ rollout guidance starts with mapping the existing setup and defining the project scope. Use your inventory to decide whether the project covers one property, a region or a wider portfolio.

You may be able to keep some hardware. ButterflyMX describes retaining existing equipment in some projects in its unified access control guide. Its access control product page also describes connections to compatible electronic or magnetic lock arrangements. These claims do not establish that every existing reader or controller is compatible.

Have a qualified installer document each lock, reader, controller, wiring run and network component that can stay, what must change and why. The assessment should map power and network needs against your required outage behavior. You need a clear project scope, not instructions to install or test the equipment yourself.

Request a phased transition plan covering temporary use of old and new credentials where supported, staff and resident training, exception handling and staged expansion. Coordinate emergency responder access with local fire officials and qualified project personnel. Emergency-access requirements vary by jurisdiction; general vendor guidance does not establish local code compliance.

If your scope includes future growth, review planning to add doors and buildings over time. First, settle what you will retain or replace. A quote based on hardware reuse is not directly comparable with one that includes replacement.

 

Compare upfront and recurring costs

Centralized access control cost depends on the number of controlled entry points, credential choice, retrofit versus new construction, hosting and service responsibility. Kastle identifies these cost variables rather than a single standard project price. Different reuse assumptions change both equipment and installation costs before you reach the total.

 

Equipment and installation

Ask for itemized quotes covering controllers, readers, initial credentials, required lock or wiring changes, network and power work, installation and commissioning. Commissioning means checking that the installed system works as required. Not every project needs every item, so have each provider mark included work, exclusions and tasks assigned to someone else.

Suppose one proposal includes installation and commissioning while another lists equipment alone. The lower total does not yet show a better deal. Ask the second provider to price the missing work and identify who will do it. You can then compare quotes covering the same obligations.

 

Software, credentials and ongoing service

Cloud versus on-premise hosting and managed versus self-managed service change what you are buying and who maintains it. Require proposals to list recurring software fees, credential issuance or replacement, applicable integration charges, maintenance and support. Record the billing unit, what each charge includes and which tasks remain with your staff.

Before comparing totals, align entry points, credential quantities, retained hardware, integrations, service coverage and the planning period. This gives you a fairer comparison, not proof that centralization saves money. Use the pilot to measure administrative work against your current process.

 

Approve a bounded pilot before expansion

A bounded pilot gives you a defined scope for checking whether the proposed administration and door behavior meet your requirements. The approach below builds on iApartments’ rollout guidance. Use it as an evaluation plan, not a promise of improvement.

 

Select conditions that represent the proposed scope

Choose entry points or properties that reflect meaningful differences in connectivity, access schedules, credentials and existing hardware. Do not choose only the easiest doors. The right pilot size depends on the conditions you need to evaluate, rather than a fixed property count.

Before configuration, define roles, who can approve changes, local exceptions and who handles unresolved issues. That gives your team a shared basis for deciding whether the pilot meets its goals.

 

Require evidence of entry, changes and recovery

The offline distinctions documented by Verkada and Kisi provide useful questions for your acceptance record:

  • Do existing and newly issued credentials behave as required?
  • When does revocation take effect, and how is it confirmed at each affected entry point?
  • Which internet, local-network and platform connections are required?
  • What power arrangements are required?
  • How are stored events recovered and made available in the shared platform?
  • Who handles unresolved exceptions?

Have qualified personnel plan and validate the relevant outage and recovery conditions while protecting continued access and emergency access. Do not disconnect operating equipment yourself to check an offline claim.

Require results for the proposed model, credential and configuration. Label observed pilot behavior, vendor statements and unanswered questions separately. A successful offline entry does not resolve an unverified revocation.

 

Measure the work against the current process

Use the same definitions for your baseline and pilot so a change in workload or observation period does not look like better performance. These are proposed measurements, not industry benchmarks:

  • Staff administration minutes per completed move-in: define the same tasks in both periods, then divide total staff minutes by completed move-ins.
  • Time to verified revocation: measure from the authorized removal request to confirmation at every affected entry point, recording the connectivity state. Keep unverified cases open rather than counting the platform action as completion.
  • Lockout frequency: define a lockout, then compare counts using the same resident-population basis and observation period.

Compare the results with your acceptance requirements, unresolved exceptions, transition burden and aligned costs. Expand only where documented behavior meets your needs. Otherwise, revise the scope, keep an accountable hybrid exception or retain local administration.

For a product-evaluation next step, explore the ButterflyMX access control system and discuss your property’s requirements. Apply the same acceptance criteria to the proposed configuration.

 

Frequently asked questions

 

Is centralized access control the same as one app for residents?

No. A shared resident credential or app describes the resident experience. Centralized administration describes how staff manage permissions. They can coexist, but one resident app does not prove that staff manage all permissions through one platform.

 

Does centralized access control require a central property management team?

No. Shared software does not require a central office or remove delegated local authority. You can organize responsibility at a property, regional or portfolio level based on the work your team needs to perform.

 

Does a controller reboot erase offline access rules?

Not in Verkada’s documented AC41 example: its last received configuration survives a reboot. Get a separate documented answer for any other proposed controller. Stored rules do not mean access continues while equipment has no power.

 

Can a rarely used credential fail even when a system supports offline access?

Yes, if the offline cache does not include that credential. Kisi documents a usage-based cache that may omit rarely used or recently added credentials. That is a product-specific example, not a universal limitation. Check user coverage alongside duration.

 

Will access events appear immediately after a disconnected door is used?

Not necessarily. Local recording and availability in the shared platform are different. Verkada’s AC41 documentation describes stored-event upload after reconnection. Ask for the proposed configuration’s recovery behavior rather than assuming immediate visibility or a universal retention period.

 

Centralize administration where shared oversight and documented door behavior meet your operating needs. Keep a justified local system or hybrid exception where they do not.

Before expanding, require evidence of both continued entry and delivery of permission changes. A door opening offline answers only the first question. Use your pilot to verify the rest and assign clear responsibility for any unresolved exceptions.

butterflymx blog header image

Get your free quote!

Fill in the form below, and we'll email you right back.

Want a free quote?

Fill in the form below, and we'll email you right back.

Content Writer
Nick has been a Content Writer at ButterflyMX since 2022. With extensive experience writing SEO-focused content as a copywriter, content operations specialist, and marketing writer, Nick appreciates the time it takes to gather and utilize data to create useful content for unique audiences.

He received his Bachelor of Arts in English from the University of Rhode Island in 2015 and his Master of Fine Arts in Creative and Professional Writing from Western Connecticut in 2017. His debut crime novel, Moon, Regardless, was published in 2021 by World Castle Publishing and his short fiction has appeared in World Unknown Review Volume III, Mother’s Revenge, SciFI Monkey’s Seasons, Death and Decorations, and more.

Nick lives in Rhode Island with his wife and son.