When we changed maintenance systems at Northpoint, the new way made sense to me. I had not spent years doing the work in the old system. For the team, a familiar task now took a different path, and I had underestimated how much guidance that would require.
That experience made me wary of two quick explanations. “People are resisting” ignores what they are trying to do. “The product cannot do it” may be equally wrong. An operator needs to look at the task and its result before deciding what the friction means.
An unfamiliar action is one possibility. The result is available, but the person has to learn where to start and what to stop doing. In our maintenance migration, orders related to a property turn could be created in the new system. Creating them outside the intended turn process, however, left the work outside the turn tracking we needed. A task guide and practice with the actual sequence are reasonable responses. The check is whether a daily user can complete the task, see the right result, and handle a problem with the guide and support they will normally have.
A missing instruction is different. A setting, default, handoff, or small choice may change the result without looking important on the screen. More general training will not fix an instruction that nobody has written down. Find the missed choice, correct the guide or configuration, and check the resulting record. Ask how earlier records will be found and corrected, too.
Then there is an outcome the current approach cannot deliver. Confirm the need, the permissions and configuration, and what an experienced user can actually demonstrate. If the required result or safeguard still cannot be produced, the response may be a process change, a narrower launch, a bounded temporary method, or reconsideration of the tool. Do not call that a training issue just because more practice is cheaper than changing the plan.
My proposed review is simple: write down the task, expected result, actual result, who is affected, and what you checked. Give urgent interruptions an owner and a safe response before completing the diagnosis. Then choose one next test. If guided practice works, keep teaching and watch whether the result holds. If the task fails despite correct use, change the process or configuration and test again. If a required safeguard cannot be shown, hold the affected scope and revisit the decision.
Discomfort alone is not proof that the change was wrong. Persistent failure to deliver the result is not a price the team should be told to endure. Listen to the people doing the work, examine the evidence, and change the part the evidence points to.
For a structured investigation, use the four-gap map. To record the decision about what can launch, use the readiness review.
Supporting source
- Year One, Episode 5, Property Meld to AppFolio: The Hardest Maintenance Migration — I discuss my experience with leader assumptions, changed task paths, missing instructions, and early support. The review method above is an editorial recommendation, not a reported Northpoint procedure.