The old system is not going to run well, but it does not keep pace with the new needs and changes are not going to affect students’ access to the Internet. It is not technology migration that makes the switch impossible, and how tens of thousands of people spend every day without anyone knowing it. We have done these kinds of projects on several occasions, and the level of disturbance depends largely on pre-designs rather than on the extent of the manufacturer.
First principle: New and old parallels, no one size fits all
The most direct pit is to choose a direct cut in the middle of the night, and the old system shuts off the new system. In the event of problems with the new system, the entire school network breaks down, fires are extinguished in the middle of the night. We insist on keeping both systems running for a parallel period of time, cutting back more than the bill. A bit of trouble comes in parallel, but any anomaly returns to the old system, and students feel almost unconscious.
Second principle: Data and rules are migrated before validation
The old system's billing rules, student sizes, and special group configurations have to be moved to the new system and tested consistently. We have seen that the rule of free flow for poor people is missing after relocation, and then the students are regularly charged and complained about it. Migration is not a copy-by-article check. Rules are right, experience is wrong; rules are wrong, and it is also wrong to cut them back.
Principle 3: A buffer period for students
If the new system changes the way it uses a volume display, a charge is added, students are informed in advance. We suggest that a first round be sent to explain what's different and how old balances are connected. The buffer period is not delayed, but psychologically prepared.
Fourth Principle: No Exchange of Periods
Examination weeks, school day and start-up, no substitutions. These are sensitive web pages, and any fluctuations are magnified. We plan to avoid these windows, choose between vacations and flat peaks. The timing is as important as the choice of options. The results are wrong when doing the right thing at the wrong time.
Fifth principle: old systems retained as retreats
When the new system is online, the old system will be kept open for a while. We usually stay at least one month and we can cut it back. We just saw the school go offline after cutting the old system, so a small configuration of the new system was not available for the day.
Principle 6: Focus on the first full cycle
The first cycle of the replacement is crucial, depending on whether the bill is correct, how many students are receiving feedback and how many abnormal numbers are unstable. We're running with the tightest view in this month, and any unusual ones are dealt with. The first cycle is smooth, and the rest is basically stable.
♪ To show us how to feel ♪
The change in the system is the simplest of technology migration, which is difficult to do without a voice when people live on a daily basis. We do these projects by using parallel, migration validation, buffers, bailouts, retreats, first-cycle watchouts as six steps. To do so, the six steps can be taken quietly, and students should be online without knowing that they have a new system behind them.
Communication rhythms affect perceptions more than technology.
We'll have a communication schedule: one month in advance, one week before the change, and then follow up on feedback. Students know how to change, how to change, how to change, how to resist. Once again stable, communications are called out.
Turn this into a next experience.
Each time we finish, we will have to remix the school: which are the right scenarios, which pits were unexpected and which were suitable for a parallel period. These experiences are written into the school ' s guidebook, which is then directly reused. The replacement is not a one-time project, but rather a build-up of school capacity. It is more long-term than replacing a system.
I'm writing a new story.
Every time we change the place, we leave a story: when it's going to be parallel, what's unusual and how to back out. This is the next map. Many schools are re-released, and they lose their experience.
Use student feedback as part of acceptance.
Instead of success, students have the most say. We suggest sending an extremely simple questionnaire after switching to see if it has been affected. The recovery of the questionnaire, while simple, is the most direct evidence of quality in replacement and more credible than self-measurement by manufacturers.
It's a chance to comb.
Instead of cleaning up the provisional rules and historical patches that have been saved for years, we've seen schools replace them with a bunch of strategies that nobody remembers. The system is much faster than technology.
The end of the peace is normal.
The alternative is not a chance to be undisturbed, but to turn the six steps into normal school-like capabilities in parallel, verification, buffers, bailouts, retreats and watchouts. Next time, just go on without retraced.
We'll have to finish the replacement.
In the final step, we suggest that a closing list be drawn up: whether the old system is completely offline, whether the refund pres are filed, whether student feedback is aggregated and whether the manual of transport is updated. Once the list has been finished, the replacement is truly over, not when the system runs. The closing list is key to turning the project into an asset.
The most dignified way to change things quietly.
We're looking for a situation where students never noticed the system changing. This is not by some chance, but by every step of it. A decent replacement is technological maturity and management maturity.