In a previous post we covered 5 best practices in a system integration project. Here we will cover 6 more best practices which are:
- Build a complete data mapping document
- Get approval on the business requirements and mappings
- Develop integration to the specifications
- Create test cases and validation plans
- Manage testing process and shared documents
- Move to production and beyond
Let’s go into more detail on these best practices:
- Build the complete data mapping – Use your initial mapping to complete your data mapping. This is an on-going process. Document intermediate mappings. Use data definitions where possible. Don’t forget to use reference data lists for mappings. Have a line by line sign off on the data mappings.
- Get approval on requirements and mappings – Now that you have documented your mappings and project requirements, get approval on these from the appropriate individuals. Get their feedback and update the documentation.
- Develop integration to the specifications – In the selected tools build the integration solution based on the approved specifications. Expect changes and have a plan to handle them (propose, review and approve). Begin working on test plans in parallel with development.
- Create test cases and validation plans – While developing, build plans and test data based on specifications. Ideally, have test instances of each endpoint that are as close to current as possible. Also make sure you have a plan to reset and rerun test data. Remember that your goal is to find problems. Better to find them now than after go-live.
- Manage testing process and shared document – Document the testing and make sure that this documentation can be shared, is accessible, can be edited and is versioned. The testing documentation should include the following:
- Testing status for all test cases
- Issue logging
- Issue resolution
- Mapping updates
- Requirements updates
- Project plan
Make sure that you have a process so everyone involved knows when changes are made. Ideally, once all test cases are approved it is ready for integration go-live.
- Move to production and beyond – Make sure that you have communicated to everyone affected the go-live date and time as well as your expectations of the affected staff. Maybe you need them to be off the system for a period. Provide the necessary staff training if there is any due to the integration project. Provide a way for errors and issues to be reported. Create automatic error notifications if possible. Create tasks for monitoring the integration. Make sure there is an assigned person to monitor the integration on a regular basis. Realize that there will be upgrades in the future that will affect the integration. If that is the case, then another integration project will need to be executed. Make sure that the integration documentation is available for future reference.
As mentioned before, integration projects are not easy. A lot can go wrong. That is why system integration project best practices need to be followed. If you have not done so, please check out the previous post with the initial best practices. Check out our video on these best practices. Also check out our other integration resources in this blog post. Get additional resources regarding linking data governance with technology projects in this blog post. Have fun on your integration project.
IData is expert in integrating higher education data systems including ERPs, SISs, CRMs and financial solutions. We have connectors and our IDataHub enterprise service bus that makes integration easy. IData also has experts that can assist with data governance, data intelligence, reporting and other technology services on an as needed basis. Feel free to contact us and let us know how we can assist.
(image credit StockSnap_6DB609DF59_Integration_BP-1 #1089)