
Discovery gives you an inventory. It tells you what exists, where it lives, who created it, and, if discovery goes deep enough, how different components relate to one another.
That is essential information for any Power Platform migration, but it does not tell you everything you need to know. As I discussed in Why Discovery Doesn’t Equal Power Platform Migration Readiness, discovery is only the starting point. The next challenge is understanding what actually matters to the business.
Imagine completing discovery across a tenant and presenting the results back to the business. You have identified several hundred Power Apps, thousands of Power Automate flows, and dozens of environments. It is a significant body of information and represents an important step forward in understanding the Power Platform estate. But then someone asks the obvious question: “Which ones are important to the business?” Unfortunately, there is rarely a convenient property that says Business Critical = Yes. That is where discovery starts to become migration readiness.
Even in organizations with strong governance, the technical information associated with a Power Platform asset does not necessarily tell you how dependent the business is on it. Understanding this requires a combination of what the technology tells you and what the business tells you – distinction matters.
Usage does not equal business criticality
One of the first places migration teams naturally look is usage. If an application has hundreds of active users, it is reasonable to assume that it is important. If an application has only a handful of users or a flow runs infrequently, it is tempting to assign it a lower priority. Usage is a useful signal, but it should not be confused with business criticality.
Consider an application that employees use every day to order lunch. It could have hundreds of regular users and generate a significant amount of activity. Now compare that with a flow that runs once a month as part of a financial close process. Technically, the lunch application may look more active. From a business perspective, the financial process may be significantly more important.
The same applies to applications with small user populations. A power application used by five people in finance or HR could support a far more critical process than an application used by 500 employees for an internal survey.
This becomes even more important with automation. Some of the most important Power Platform assets may have no interactive users at all. They simply operate quietly in the background and drive results.
Start with the business impact
One of the simplest questions to ask about an application or flow is also one of the most useful: What happens if this stops working tomorrow? It changes the conversation from technical activity to business impact.
As an example, if an office lunch ordering application stops working, it will certainly inconvenience people. If an internal department survey is temporarily unavailable, somebody may need to delay sending a survey. Those applications still have value, but the business can probably continue operating while the issue is resolved.
Now consider an application that supports customer onboarding, employee onboarding, invoice approvals, purchasing, or compliance reporting. If one of those processes stops, the consequences can be very different. Employees may be unable to complete their work. Payments could be delayed. Customers may be affected. A compliance activity could be missed.
That is really what migration teams are trying to understand when identifying critical assets. It is not simply whether an application is popular or heavily used. It is what happens to the business when it is unavailable. Such information should influence how the asset is treated during migration. A business-critical solution may need more detailed dependency analysis, more testing, a clearly identified business owner, a controlled cutover, and somebody explicitly responsible for validating it in the target environment.
The challenge is knowing which power app or power automate flow requires this level of attention.
Ownership tells its own story
Through discovery we now understand who created an app or flow, however, but that can raise as many questions as it answers.
Does the owner still work for the organization? Are they actually responsible for the business process, or did they simply build the solution? Is there somebody else who understands how it works? Who supports it today? And, importantly for migration, who is going to confirm that it still works afterwards?
This is where migration teams often find uncomfortable situations. An application may be owned by a former employee. A flow may be associated with an account that is no longer actively managed. Nobody may be entirely sure what a solution does, and finding someone prepared to validate it after migration can be surprisingly difficult. It is tempting to look at those assets and assume they are probably unimportant, which can be a dangerous assumption to make.
A three-year-old flow created by somebody who left the company two years ago may still be moving information between systems every day. An application without an obvious business owner might still be supporting an important departmental process. Nobody knowing what something does is not evidence that nobody needs it. If anything, it can represent additional migration risk because the organization has lost the knowledge required to understand, support, and validate that solution.
Looking beyond the app itself
Another way of determining criticality or importance is understanding dependencies.
A Power App may look relatively straightforward when viewed in isolation, but the application itself may only be the visible front end of a much larger business process.
For example, a user might submit information through a Power App. That action could trigger Power Automate, update data in SharePoint or Dataverse, send an approval through Teams, interact with another system through a connector, and ultimately feed information into a Power BI report.
The application might be simple. The process behind it is not. This is why dependency mapping becomes so important during migration planning.
If you only look at individual assets, you risk treating every Power App and flow as a separate migration object. The business is not thinking about them that way. The business is thinking about the process they collectively support. A broken connector, missing dependency, identity change, or configuration issue in one component may affect the entire process. In some cases, the dependency map tells you more about the importance and migration complexity of an asset than the app itself.
Don’t forget the flows nobody talks about
Interactive applications are visible. People know when they use them, and generally somebody notices fairly quickly when they stop working. Background automation can be very different.
Think about flows that quietly route invoice approvals, synchronize data, provision users, send HR notifications, produce compliance information, or move information between business systems. Nobody opens those flows every morning. Most users probably do not even know they exist. As long as they continue to work, nobody talks about them. But once they stop working, everybody notices.
These are the silent dependencies that can easily be missed if migration prioritization is based primarily on visible usage or user counts. A flow that runs once a week or once a month may be supporting something far more important than an application used hundreds of times each day.
That is why migration teams need to look at what an automation does, what it connects to, and what business process depends on it, rather than simply how often it runs.
Five questions to ask
So how do you identify what is business critical? A complicated formula is not needed. The technical information gathered during discovery should be used to focus the conversation with application owners, process owners, and the wider business. For each solution, I would start with five questions:
- Would business operations be disrupted if this stopped working?
- Does it support a revenue, finance, HR, compliance, or customer-facing process?
- Does it connect multiple systems, services, or data sources?
- Does it require specific business validation before or immediately after migration cutover?
- Would the business immediately contact IT if it failed?
Instead of treating discovery as the end result, use the inventory as evidence to ask better questions. The objective is not to invent a perfect criticality score, but to distinguish between “we discovered this” and “we need to make sure this continues working.”
Technical discovery tells you what exists. Business context tells you what matters. Migration readiness comes from bringing those two things together.
When planning a migration, finding hundreds of Power Apps and thousands of Power Automate Flows is useful. Knowing the ones that the business depends on is where discovery ends and migration planning begins.
