What Do I Need to Manage Multiple Sites in One Application?
By Adrian Butnaru · Independent developer · Blanes (Girona), working remotely across Spain.
Adrian's GitHub · Adrian's LinkedIn.
Published: · Last technical review:
A multi-site web application requires defining which technical records and incidents belong to each location, which users access a single site or multiple sites through roles and permissions, and which PDF reports or push notifications are managed per site. I analyse your organisation and we agree on that scope in writing before development.
The site name does not settle the requirement
If you tell me that you need an application for multiple branches, I would ask you to start with the work that takes place at each one. Do they collect the same records? Are there people working across different sites? Do you need to view each site separately or prepare a combined overview? I prefer to organise these decisions before talking about screens.
I would not treat «multi-site» as a complete answer. In my published offering, that capability appears alongside roles and permissions, push notifications, and PDF export. What remains to be defined for your case is how you want to organise the information and what actions each user profile needs to carry out.
A map to explain your organisation
You can prepare this inventory:
- What sites should appear in the application.
- What records correspond to each site.
- Which people need to act across more than one location.
- What documents you want to consult per site.
- What information you would like to bring together in a consolidated view.
- Who will decide the rules when an exception arises.
There is no need to send me personal data to get started. You can describe fictional sites and profiles as examples of the requirement, always identifying them as such. What I want to understand is the relationship between a task, its information, and the site it belongs to.
Local data and shared requirements
I suggest separating what should remain associated with a specific site from what you might want to reuse. A list of record types, for example, could be approached as a shared requirement; the content of an individual record would require identifying its context. I present this as a design question, not as a description of how your future application works.
I also need to know what «seeing everything» means to you. It might refer to consulting sites separately, comparing information, or compiling it into a single document. I would not assume these requests are equivalent, nor that they are included simply by mentioning an overview. I would ask you to describe what decision you want to make with that information.
Editorial framework, not implemented architecture
Editorial framework for a site-specific workflow
- Input: selected site, profile, and record they need to use.
- Proposed validation: verify whether the person has permission for that site and action.
- Action: view or modify only the information agreed upon for that workflow.
- Output: prepare the PDF report or query for the requested site.
- Exception: if the profile lacks access, define how to explain the restriction without displaying other parties' data.
Consolidated query branch
- Input: purpose of the summary and sites included.
- Proposed validation: agree on who can view that combined set.
- Output: distinguish records by site without losing their origin.
- Exception: if permission is missing for one site, specify whether it is excluded from the summary or if generation is blocked; do not decide by default.
This framework serves to review scope. It does not demonstrate technical isolation, database separation, or a specific implementation. Those assertions require specific evidence, and I do not deduce them from a screenshot or the commercial name of a feature.
How to describe a workflow across sites
I propose writing one local workflow and one shared workflow without attributing them to an existing application. In the local workflow, identify who records information, which site it belongs to, and who needs to review it. In the shared workflow, explain which person needs to consult several sites and for what task. These are proposals to prepare requirements, not examples of an already delivered implementation.
A combined query also needs exclusions. Note down what should happen if someone can work at one site but not another, or if their responsibilities change. I would not choose an answer by intuition: the decision affects what that person expects to see and what the project will need to verify.
You can describe these scenarios using fictional site and profile names, clearly identified as examples. I do not need real records to understand the requirement. If reviewing documentation becomes necessary later, we would need to agree on what material is required and how to share it.
Before translating this framework into a quote, I suggest distinguishing what has been decided from what remains open. Having a shared application does not on its own resolve who administers each site or how an exception is authorised. The framework guides the definition of requirements; it does not constitute a closed scope, nor does it substantiate the rules of an existing application.
When you should NOT bring everything together
I would not suggest centralising information merely to reduce the number of tools. First, I want to know whether the sites truly share the process and whether you need common coordination. If the tasks are different, we may need to define separate workflows rather than imposing an identical form on everyone.
As a comparison, I would weigh separate tools against a shared application. Separate tools allow discussing each process on its own; a shared application requires explicitly agreeing on common ground and its boundaries. I do not claim that one alternative is superior or that your organisation will save with it.
Nor would I accept «letting everyone in» as sufficient criteria. I need to distinguish who can access the system and what they should be able to do once inside. If that is not yet decided, I would leave it as pending, rather than presenting an improvised structure as a closed solution.
How I work through this definition with you
To assess your multi-site application, I review one local workflow and one consolidated query requirement. With that information, I prepare the written diagnosis and agree with you on which sites, views, and documents form part of the development, without turning one site's requirement into universal permission for all of them.
To prepare that conversation, I suggest selecting a representative workflow: one person views or logs information at a site and another needs to review it. Then add a scenario that should remain restricted. That way, we can discuss both the standard task and the boundary, without attributing capabilities that have not yet been agreed upon.
What I want to distinguish in the quote
I would ask you to specify whether you require views per site, user access management, document exports, or a consolidated query. I would not group them under a single phrase that seems to include every variation. I would also note down what is excluded and what would need further analysis before committing to it.
If you hold data in previous tools, I want to know from the start. I do not take a migration or an integration as included. You can describe what exists and what you want to preserve; afterwards, we would need to review the specific scope rather than assuming that centralising equates to transferring everything.
What to prepare before writing to me
In the checklist self-assessment, you can specify these three points: whether your operations cover multiple sites, who views or modifies data at each site through roles and permissions, and what push notifications and PDF reports you require per site.
If an answer is not yet clear, you can leave it as an open question to review. What is useful is identifying the shared workflow and the most important exception. With that, I can evaluate the starting point with you without turning the checklist guidance into a technical diagnosis or an implementation commitment.
What information must preserve site context
I suggest checking whether your description makes it possible to recognise which site each data point belongs to, who needs to consult it, and what exceptions must be discussed. This is a requirements exercise, not a demonstration of technical data isolation or a behaviour attributed to an existing application. If a decision is missing, it is best to keep it visible before closing the scope.
For your project, you can describe a record that belongs to a site and a person who needs to review it. Then add what changes when that person works at another site: the data, the authorisation, or the requested document. If you want a combined view, specify its purpose and the sites included. The act of centralising does not on its own resolve what information can be shared, nor what exceptions must remain restricted.
Frequently asked questions
What information belongs to each site?
Information in a multi-site application must be linked to the site it belongs to: records, incidents, and documents need that context. I review with you what data is local, what you want to share, and who will use each view. Multi-site management and its boundaries are agreed upon in writing, not deduced from the service name.
How do I organise user access?
User access in a multi-site application is organised by profile, action, and site. I distinguish with you who views or modifies data at one site and who needs to work across several. I suggest putting exceptions and boundaries in writing, without deducing them from a screenshot or the commercial name of a feature.