According to PMI, which of the following scenarios is best and most responsible?
Context: For the Golive of a PO creation software implementation project, the total number of users who create POs exceeds 50 people and affects 3 countries, it is recommended.
1. Go to GoLive with 10% of the users during the first month and release the remaining 90% according to results.
2. Going out to GoLive with 100% of the users. Saving Changes...
There is not enough information provided to say. The question comes down to what is the risk of finding issues as you implement your system that might be mitigated through a limited implementation?
50 people is a small user base, but if issues could affect all of your purchasing, that might still cause a major supply chain disruption. Testing it out with a limited number of users can limit the disruption from problems encountered and allow early users to train others for full implementation. That can significantly extend your timeline and force you to operate in multiple parallel systems which has risks and costs of its own.
I would suggest showing the plan for both scenarios, evaluating the pros and cons, and making your decisions off the analysis. What someone at PMI thinks is "best" based on a generic situation will not help you.
In addition to Keith's feedback, I'd suggest that constraints might also affect the decision. For example, if managing two different types of POs for a period of time will create too much additional work or introduce too much risk, a big bang approach might be necessary.
There might be a contractual obligation as well which requires a full go live.
There is not enough information provided to say. The question comes down to what is the risk of finding issues as you implement your system that might be mitigated through a limited implementation?
50 people is a small user base, but if issues could affect all of your purchasing, that might still cause a major supply chain disruption. Testing it out with a limited number of users can limit the disruption from problems encountered and allow early users to train others for full implementation. That can significantly extend your timeline and force you to operate in multiple parallel systems which has risks and costs of its own.
I would suggest showing the plan for both scenarios, evaluating the pros and cons, and making your decisions off the analysis. What someone at PMI thinks is "best" based on a generic situation will not help you.
In addition to Keith's feedback, I'd suggest that constraints might also affect the decision. For example, if managing two different types of POs for a period of time will create too much additional work or introduce too much risk, a big bang approach might be necessary.
There might be a contractual obligation as well which requires a full go live.
As always, it depends.
Kiron
Thank you very much Kiron. Saving Changes...
Thomas WalentaGlobal Project Economy ExpertHackenheim, Germany
Jose,
as the others say, it depends.
I managed many GoLives with SAP, which includes PO (purchase orders?) but has a much wider functional range.
I always did a big bang GoLive with all users. Maintaining 2 productive databases offers a bunch of risk in itself (data integrity, 2 applications/processes, how to reconcile). We tended to do it at (financial) yearend to start with a clean slate of transaction for the new year / new application.
We sometimes did prototypes with a small population, but not for productive use. As a project you can play around, but in production you better make sure not to hamper with business.
If, as a scenario, your complete user base is not ready, do not do the GoLive… The problem of incomplete change management should not be solved by slicing data and processes.
Thomas
...
1 reply by Jose Chaguala Garzón
Dec 22, 2022 10:24 AM
Jose Chaguala Garzón
...
Thank you very much Thomas, and yes, PO is purchase orders.
Saving Changes...
Sergio Luis ConteHelping to create solutions for everyone| Worldwide based OrganizationsBuenos Aires, Argentina
In my actual work place we call Go Live to move to production the final product. After that we perform some validation and then we Relase the final producto to end users. So, we perform Go Live in a controlled way with key users whom performed the validation just in case we have to do a rollback. Saving Changes...
Stéphane ParentSelf Employed / Semi-retired| Leader MakerPrince Edward Island, Canada
There is value in doing a parallel implementation, keeping the old system around while your stakeholders get working with the new one. Parallel systems give you the option to go big bang or progressive deployment.
The drawback with progressive deployment is the need to repeatedly bring data from the old system into the new system. Of course, if you have a successful data conversion process, there's nothing stopping you from scheduling it as a regular job. Saving Changes...
Community Champion
