In my last post, Project Online Retirement: How to Choose What Comes Next, I walked through how one organization chose its path off Project Online. This time I want to talk about the part of a migration that almost nobody plans for.
Everyone plans to move the projects. Very few organizations plan for what has been quietly running around them for years. In my experience, that is where most of the risk in a Project Online migration actually lives.
Over time, a project management platform collects connections. A Power Automate flow keeps resource rates up to date. Another sets up new users when they join. A report pulls numbers from the finance system. A macro somebody wrote years ago fills in fields every time a plan is closed.
Each of these made sense when it was built. The trouble is they were usually built by different people, at different times, and very often nobody owns the full picture. When the platform underneath them goes away, so does everything they were quietly holding together.
The client from my last post is an engineering and testing organization moving off Project Online. On paper, they had a list of about 20 automated workflows. In practice, those workflows were keeping resource rates, user setup, access, and reporting in sync every single day, and the people relying on them mostly saw the results, not the automation.
Before we rebuilt anything, we did one thing first: an assessment. For every workflow, we asked what business problem it was solving, whether that problem would still exist on the new platform, and what could go wrong if it simply stopped.
The natural instinct is to treat a Project Online migration as a copy job: list the flows, rebuild them one for one on the new platform, and move on. That misses two important points.
First, how you rebuild depends entirely on where you land. This client chose Project Server Subscription Edition on their own servers, so their automation became scheduled jobs running alongside it. If they had moved to a cloud platform, the same needs would have been met with Power Automate or the platform’s own automation features. A different PPM product, like OnePlan, might handle some of it out of the box.
Second, some of those problems only existed because of Project Online. Move to a different platform and the problem may simply disappear. Rebuild the automation anyway, and you are carrying forward a fix for something you no longer have.
That is why the assessment matters more than the rebuild. It tells you what to keep, what to retire, what to combine, and what to rethink before you spend a dollar on building anything.
One workflow ran eleven times a day. It existed to repair a field that Project Online’s own directory synchronization kept clearing. On the new platform, that problem did not show up in testing, so the job now runs once a day as a safety net instead of eleven times as a repair crew.
The lesson: some automation exists to fix problems you may not have anymore.
Two of the workflows were not worth rebuilding at all. One had never been run in three years. The other turned out to be broken, and a faithful rebuild would have written blank values into live data. We confirmed both with the client and documented why they were left out, so nobody “fixes” the gap later.
The lesson: don’t port blindly, and write down why something was left behind.
The workflow that deactivated departed employees had no safety check. If the directory lookup ever came back incomplete, it could have deactivated almost everyone in a single run. It also caught current contractors who were simply missing one HR attribute. The rebuilt version refuses to act on a suspicious result and reports it to a person instead.
The lesson: automation that removes things deserves the most care, wherever it runs.
While rebuilding the reports, we found that the client’s finance system was relying on internal record IDs from Project Online. Those IDs change during a migration. Nobody had documented the connection, and nobody on the project team knew it existed. Finding it before go-live meant fixing it on our schedule instead of discovering it when the numbers stopped matching.
The lesson: the scariest dependencies are the ones that were never written down.
In the end, about 20 workflows became 8 scheduled jobs. Each one is documented, does one clear job, logs what it did, and emails a short summary when something needs a person’s attention. The business outcome stayed the same: rates are right, users are set up, access is current, and reports match. There are simply fewer moving parts, and the client’s team knows what each one does.
If they had landed somewhere else, the tools would look different. The decisions would have been made the same way.
Whatever platform you are moving to, a useful assessment answers the same questions for every automated process and integration:
The output is a short list of decisions your team can stand behind, not just an inventory.
At Pinnacle Management Systems, we start every Project Online migration with this assessment, and we can run it on its own if you are still deciding where to go:
Do Power Automate flows keep working after migrating off Project Online?
Flows that connect to Project Online need to be rebuilt or retired, because the service they talk to no longer exists. How you rebuild them depends on your new platform: a cloud platform usually means Power Automate or built-in automation, while an on-premises Project Server usually means scheduled jobs on your own servers.
Should we rebuild every workflow we had in Project Online?
No. Some solve problems that won’t exist on your new platform, and some may not be used at all. An assessment tells you which ones are worth rebuilding before you spend time and budget on them.
When should the assessment happen?
Before you commit to a platform, if you can. Knowing what is connected to your current environment helps you choose where to go. For this client, the assessment came early, before any projects were moved.
We already migrated. Is an assessment still worth it?
Yes. Automation that silently stopped working after a migration is easy to miss until a report is wrong or someone can’t log in. A quick review can confirm everything that should be running still is.
Does this apply if we are moving to OnePlan or another PPM platform?
Absolutely. The tools change from platform to platform, but the questions are the same: what does each piece of automation do, does the business still need it, and what is the right way to deliver it on the new platform.
If your organization is planning a move off Project Online, or has already moved and isn’t sure everything came along, I’m happy to talk it through. The assessment is usually where the most valuable conversations happen.