Power Platform Migration: Finding the Assets the Business Actually Depends On
Discovery shows what exists in your Power Platform estate. Migration readiness depends on knowing which apps and flows the business can't afford to lose.

Discovery gives you an inventory. It shows what exists, where it lives, who built it, and, if you dig deep enough, how the components connect. That's essential for any Power Platform migration, but it still leaves the hardest question unanswered. Which of these things actually matter to the business?

Picture presenting discovery results for a tenant: several hundred Power Apps, thousands of Power Automate flows, and dozens of environments. It's a substantial body of information and a real step forward. Then someone asks the obvious question, and the answer is rarely sitting in a convenient field. Nobody tags assets with Business Critical = Yes. That gap is where discovery turns into migration readiness.
Even well-governed tenants don't reliably tell you how dependent the business is on a given asset. Technical metadata describes the asset. Business context describes the consequences of losing it. You need both, and keeping them separate in your head is the first discipline to build.
Usage is a signal, not a verdict
Migration teams often reach for usage first. An app with hundreds of active users feels important, and one with a handful of users or a flow that runs rarely feels like it can wait. Usage is useful evidence, but it isn't the same thing as business criticality.
Take an app employees use daily to order lunch. It could have hundreds of regular users and plenty of activity. Now compare a flow that runs once a month as part of a financial close. The lunch app looks busier on paper. The financial process is almost certainly more important to the business.
Small user populations can hide critical work too. A Power App used by five people in finance or HR might support a far more consequential process than one that 500 employees use for an internal survey. And automation makes this worse. Some of the most important assets in your estate may have no interactive users at all. They run in the background and quietly drive outcomes.
Ask what happens when it stops
One of the most useful questions you can put to an app or flow is also one of the simplest: what happens if this stops working tomorrow? That reframes the discussion from technical activity to business impact.
If the lunch ordering app goes down, people will be inconvenienced. If an internal survey is unavailable, someone may have to delay sending it. The business keeps running in both cases. Now consider an asset supporting customer onboarding, employee onboarding, invoice approvals, purchasing, or compliance reporting. A stoppage there can stall staff, delay payments, affect customers, or cause a compliance activity to be missed.
That's what migration teams are really trying to establish. Impact should shape how an asset is handled during the move. A business-critical solution will usually justify deeper dependency analysis, more thorough testing, a named business owner, a controlled cutover, and a specific person accountable for validating it in the target environment. The hard part is identifying which apps and flows deserve that treatment in the first place.
Ownership raises as many questions as it answers
Discovery tells you who created an app or flow. That's useful, but it often opens more questions than it closes. Does the owner still work for the organisation? Are they responsible for the business process, or did they just build the thing? Does anyone else understand how it works? Who supports it today? And who will confirm it still works after migration?
This is where things get uncomfortable. An app may belong to a former employee. A flow may be tied to an account nobody actively manages. Nobody may be certain what a solution does, and finding someone willing to validate it afterwards can be surprisingly hard. It's tempting to assume those assets are unimportant. That assumption can be dangerous.
A three-year-old flow built by someone who left two years ago may still be moving information between systems every day. An app with no obvious owner may be propping up an important departmental process. Nobody knowing what something does isn't evidence that nobody needs it. If anything, it signals added migration risk, because the organisation has lost the knowledge needed to understand, support and validate the solution.
Dependencies reveal the real process
Criticality also depends on what an asset connects to. A Power App can look straightforward in isolation while being only the visible front end of a much larger process.
Consider a user submitting a form through a Power App. That single action might trigger a Power Automate flow, update data in SharePoint or Dataverse, send an approval through Teams, call another system through a connector, and feed a Power BI report. The app is simple. The process behind it isn't.
Look only at individual assets and you risk treating each app and flow as a separate migration object. The business doesn't see them that way. It sees the process they support together. A broken connector, a missing dependency, an identity change or a configuration issue in one component can take down the whole chain. Sometimes the dependency map tells you more about an asset's importance and migration complexity than the app itself ever will.
The flows nobody talks about
Interactive apps are visible. People know when they use them, and someone usually notices quickly when they break. Background automation behaves differently.
Think about flows that route invoice approvals, synchronise data, provision users, send HR notifications, generate compliance information, or shuttle information between business systems. Nobody opens those flows each morning, and most users probably don't know they exist. While they work, nobody mentions them. Once they stop, everybody notices.
These silent dependencies are easy to miss if prioritisation relies mainly on visible usage or user counts. A flow that runs weekly or monthly may support something far more important than an app used hundreds of times a day. So look at what an automation does, what it connects to, and which business process depends on it, rather than how often it fires.
Five questions to put to owners
You don't need a complicated scoring formula. Use the discovery data to focus conversations with application owners, process owners and the wider business. For each solution, 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?
The aim isn't to invent a perfect criticality score. It's to move from "we discovered this" to "we need to make sure this keeps working." Treat the inventory as evidence that sharpens those questions.
Technical discovery tells you what exists. Business context tells you what matters. Migration readiness comes from bringing those two together.
Finding hundreds of Power Apps and thousands of flows is useful. But discovery only ends, and migration planning only begins, once you know which of them the business can't do without.