12 Apr Migrating HR Data: Best Practices
Written by: Sanjeev Singla – Software Development Lead, HRsoft
In case you missed the first article in this two part series, (Migrating HR Data: The Challenges) I recently had the opportunity to manage a data migration project for a new customer who chose one of our many talent management solutions. I would now like to share a few of the best practices that our team used to execute and implement a new HR system that fit the specific needs of our customer.
- Understanding The Data Model
First, our team divided the work of this project into three categories: business understanding, the mapping of the new data and the actual import of the data. The key to making this major data migration work was the collaboration between our business and professional services teams and our technical and development teams. The data mappings were prepared and reviewed internally within these teams and other knowledgeable individuals.
While I don’t want to go into too much technical detail, I would like to highlight a few aspects of our approach which worked in our favor:
- A semi technical person on the business side imported the raw data into staging tables and created the SQL views.
- The development team did the heavy lifting in terms of importing the new data to its correct place with all the relationships, constraints and validations.
- These two teams worked independently without slowing each one down in terms of finding and mapping new data and importing the various elements into the application.
- Parallel & Agile Approach
Throughout this project, our various teams worked on the multiple data elements import in parallel. As the elements or group of data elements were being found and mapped, the import was started in parallel for initial validation. We moved things quickly from DEV, QA, and UAT to LIVE. Everything was also moved incrementally from DEV/QA to UAT so it could be tested as it was imported.
Our team then grouped logical entities into one task. The data related to one entity was imported with all of the relationships and constraints without affecting other entities. The items could be then tested independently and in parallel.
We also accelerated the deployment by having development access to a standalone virtual machine which was not in a production pool and could be used to import the data. This cut down a lot of time throughout the normal process.
With any major migration project, prioritization is key. For this specific project, we proposed to load active in-use data first and then load the remaining data over next couple of weeks and weekends. Our customer did not have any concerns with this, as they needed the active data first to Go-Live and rest of the data to be imported within next few days/weeks.
One example of a prioritization challenge that we faced was importing attachments. We quickly assessed that the attachments could be dealt independently and the task was assigned to a completely different developer to import the attachments along with underlying data for attachments. Due to their sheer size and volume, the attachments were also incrementally uploaded in the talent management system.
- Risk Assessment, Mitigation, & Scope Changes
Our team applied their expertise and analytical techniques along with regular meetings to identify the possible risks associated with a migration of this nature. In order to avoid those risks and to plan for a smooth transition, a mitigation plan was established and ready to address any risks.
For example, we estimated the impact of a few scope change requests that arose during the course of the project and mitigated the risk by intelligently incorporating those into other data import elements.
- No Disruption to Live System
Since we are running a multi-tenant SaaS talent management application, we could not risk the live system going out of order due to massive data imports running together. To prevent this, we scheduled data import tasks over weekends and nights to minimize any disruption to the LIVE environment and closely monitored those tasks to make sure prompt action was taken if there were any issues during the data import.
The inbuilt scheduler on the platforms of our talent management applications was critical. We could schedule tasks to run at various time intervals and could specify the start and end times. All of the tasks were scheduled to run at odd times when there is less customer activity. Therefore, it didn’t interfere with the usual traffic on the HR solution and prevented any issues from arising.
- Customer Engagement
Engaging the customer is key to the success of any project. Our professional and business teams did a great job gathering feedback and keeping the customer updated throughout the various phases of the project. Early feedback and testing the data on the UAT environment helped us to fix issues and make change requests. The constant negotiations and clarifications with the customer on what data elements are key to the application functioning, also proved beneficial.
We involved the customer right from beginning. As the data elements were being imported we began sending screenshots from our Development/ QA environments to the customer. Since the customer still had access to the old system, they then could compare the data early-on.
- Effective Communication within all Teams
Like any other project we had daily sync up meetings for project success, status and the assignment of tasks. While this is a normal practice in an agile or even traditional SDLC, we did lot of review in terms of the assignment of tasks, prioritization and incremental changes to various environments.
- Process Optimization and Tailoring
We internally optimized what processes needed to be used in order to ensure a fast and error free transition for our customers, making anything related to this particular project first priority for all the teams. We saved a lot of time by tailoring the processes that were needed for this project. For Example, we did not follow the traditional ticket entry for all the issues found during data import but we used other mechanisms/tools to manage that part of the project, ultimately saving a lot of time.
Finally we made sure that the final deliverables were accepted and most importantly that our customer was happy with their new talent management system and the support that they received from our end. We updated our organizational process assets and documented all the key learnings from the project.
HRsoft is a leading provider of strategic talent management software that improves manager effectiveness and business results. Our full suite of cloud-based HR software solutions includes applicant tracking software, compensation planning software, total rewards software, stay interview software, performance management software, and content management software. To learn more about our Talent Management Software, click the button below!