- Posters in the break room
- A launch email from the CIO that nobody finishes reading
- A training video pushed out the week before go-live, watched at 1.5x with the sound off
- A completion box marked 90%, which measures attendance rather than capability
- A workstream bolted on at the end, reporting green because it has nothing to report
- A sponsor who refuses exceptions in public
- Stakeholder analysis that maps whose day changes and by how much
- Training built on your data, your processes, in the role people actually hold
- Adoption measured in the system, by transaction, by department
- A discipline that runs the length of the program and past go-live
I have watched a go-live go off without a single technical defect and still fail. The integrations held. The data reconciled. The cutover ran on schedule. Two weeks later finance was closing the month in spreadsheets, the warehouse had gone back to its old workarounds, and the help desk was drowning.
The system worked. The people were not ready.
You can spend two years and tens of millions building exactly the right system. If the people who have to use it on Monday do not trust it or do not know how, you have bought an expensive way to do what you were already doing.
1Why the checkbox version fails quietly
When change management is a checkbox it does not fail loudly. Adoption numbers look fine on paper while people route around the system in the background, and you find out when the quarter closes wrong.
Completion percentages that are high and rising while help desk volume is also high and rising. Those two numbers should move in opposite directions. When they do not, the training was consumed rather than absorbed.
2Sponsorship means more than a logo on a slide
The single biggest predictor of adoption is visible, active executive sponsorship. Active means the sponsor shows up, uses the language, makes the hard calls, and holds their own leadership accountable in their part of the business.
Change cascades from the top or it does not cascade at all. If a division head privately signals that the old way is fine for now, their entire organization hears it, and no amount of training reaches them. I have seen one skeptical executive quietly sink a deployment the project team otherwise ran well.
If the answers are vague, the sponsorship is decorative, and it is better to know that early than to find out during deployment.
3Know exactly who you are changing
You cannot prepare people you have not bothered to understand. Real stakeholder analysis is unglamorous and it is the work. Map who is affected, how much their actual day changes, and where the resistance sits, which is rarely where the org chart suggests.
4Training that sticks
Generic vendor courseware teaches the product. It does not teach your job in your system. The gap between those two is where the help desk volume comes from.
| Instead of | Do this | Why it holds |
|---|---|---|
| Generic product training | Role-based paths built on your configuration and your data | People learn the transaction they will actually run, not the demo version of it |
| One push the week before go-live | Layered delivery: awareness early, capability close in, reinforcement after | Nothing survives a single exposure two weeks before a change of habit |
| Video completion as the metric | A practical task completed in a sandbox, scored | Completion measures attendance. A completed transaction measures capability |
| Central team answering everything | A named super user per department, trained early and given standing | People ask the person two desks away before they ever open a ticket |
| Documentation as the fallback | Job aids at the point of work, in the flow of the task | Nobody reads a manual at 4pm on a Friday with a queue in front of them |
5Measure adoption, do not assume it
Adoption is observable in the system. You can see who is transacting, who is not, and who is doing it once and then going back. Watch it by department, weekly, and act on the outliers while they are still outliers.