Collaborate 15 White Paper Implementing Oracle Fusion Payroll
- 1. COLLABORATE 15 Copyright @2015 by Hitachi Consulting 1 Implementing Oracle Fusion Payroll: A Report from the Front Lines Larry Freed Overhead Door Corporation Bill Stratton Hitachi Consulting Abstract Overhead Door Corporation, along with their implementation partner, Hitachi Consulting, recently completed an implementation of Oracle Fusion Payroll on January 1, 2015. The implementation team included a unique blend of PeopleSoft, Oracle EBS and Fusion HCM experience with a combined 74 years of implementation experience. This presentation will outline our initial impressions of this module including: Things that are different: Fusion offers some new structures and ways to think about your organization. You need to fully understand these structures to leverage the new functionality. Things we like: Although Fusion Payroll looks very similar to Oracle EBS Payroll, there are some new features that we really like and wished we had them in our previous pre-Fusion implementations. Things we hope make it in the next release: With any new release of software, there are things we wished worked better than they do. We will share with you some things to look out for as you go through your implementation. This presentation will include both functional and technical observations. We worked closely with Oracle Support during the life cycle of this project, so we will share with you tips and advice on how to work efficiently with Oracle Support to quickly resolve your implementation issues. Overhead Door Background Some of the key metrics of the workforce at Overhead Door include: Approximately 3500 employees over 27 states including Ohio and Pennsylvania. These last two states have several local taxes that caused issues that will be discussed later. The scope of the payroll project only included the US legislation. Overhead Door has three legal entities in the US. We also loaded employees for Canada, UK, Mexico and China. There is a weekly payroll for all of the hourly, including union, employees. We also process a bi-weekly Non-Exempt Salary payroll and a bi-weekly Salary payroll. All three of these payrolls have separate payroll calendars. Advanced Benefits was also included with this implementation. All benefit deductions were calculated in OAB and passed directly to Payroll. We did have some minor Fast Formula modifications to support the calculation of Union Dues and Shift Premiums. We also modified the formulas for the FSA benefit deductions to self-adjust so that the employee will reach their annual goal amount. This will handle the issues when a deduction was missed or the employee changes their goal during the calendar year.
- 2. COLLABORATE 15 Copyright @2015 by Hitachi Consulting 2 Kronos was the main time management system. We developed interfaces to and from Kronos to process time. Kronos batches were loaded into the Batch Loader environment using Fast Formulas to parse the detail time records. Locations not on Kronos sent their time via spreadsheets that were eventually loaded into Fusion using the Batch Loader. First Impressions We attended Oracle training on Fusion before the project started. Greg Clark, Oracle Support Manager for Fusion Payroll, was also in the class. His remark pretty much sums up our initial impressions “It’s EBS Payroll on steroids”. Seasoned EBS payroll users will quickly recognize the features of Fusion Payroll. The question becomes not how to do something, but where is the screen? The learning curve is more on navigation than functionality. Some of the terminology is different, but the engine is very similar. Some of the challenges of navigation are whether to use the Functional Setup Manager for a particular task or whether to use the base Navigator. Several initial setups do require using the Functional Setup Manager, like setting up Legal Entities and entering the various parameters like SUI rates. Other setups can be performed from the standard Navigator, like Payroll and element setup. After struggling initially on where to perform a task, a user can quickly navigate through the system to find the desired task. What’s improved? The biggest obvious change is the absence of Business Groups. This has been replaced with the concept of Legislative Data Groups, where most of the legislative related attributes are stored. We no longer need to have multiple person records in the system for employees that have activity in multiple countries or legislations. The next new concept to grasp is Relationships. Many of the payroll attributes that were originally on the Assignment screen have now been moved to the Payroll Relationship. Now HR can make changes to a person’s work assignment (like legal entities, salary basis) and not have to change the payroll relationship. This eliminates the issue we all had on EBS Payroll when the payment date is after the period end date and we wanted to update the Assignment screen between the period end date and the payment date. EBS would always stop us since there was a payroll action after the desired change date. We also have the capability of assigning elements at the payroll relationship level instead of the assignment level. This simplifies the process when multiple assignments are being used. This is especially important for Benefit and Wage Attachment deductions. Some of the enhancements made on the HR side do trickle down to Payroll. There are an increased number of “tiers” where attributes can now be assigned. We have already discussed setting elements at the Payroll level. You can also introduce a third tier called Employment Terms that sits above the Assignment and Work level. You can use this third tier for more complex hierarchies that exist in higher education and public sector implementations. Fast Formula Enhancements For users that are Fast Formula savvy, you will see a number of enhancements to Fast Formulas. Working in the Cloud, you will no longer be able to write and catalog your own Fast Formula functions. Therefore, Oracle has provided more ways to access database items through the concept of “contexts”. For example, while processing an employee, you can set a context using the person id of an employee’s dependent or spouse. Once set, you can then access attributes associated with this contact. You can also set a context for a previous payroll run to access various dimensions (Run, YTD, MTD) of payroll balances as of that particular payroll run. The goal was to eliminate
- 3. COLLABORATE 15 Copyright @2015 by Hitachi Consulting 3 the need to write custom Fast Formula functions and procedures to access data that normally is not available through database items. Formulas in general are much more object-oriented. Most elements actually have separate formulas to load input values into arrays, perform base calculations, perform proration and other tasks that were included in the one base formula. There is now a special Formula Type called Payroll Access to HR data that helps with accessing the various data attributes that are stored in HR. Below is an example of a Fast Formula we wrote to access the Location Name, Shift indicator, and Job Indicator that was stored in one the Assignment screen and various descriptive flexfields. The first formula is the Payroll Access to HR Data type formula that retrieves the desired attributes: DEFAULT FOR PER_ASG_LOCATION_NAME IS 'NOT ENTERED' DEFAULT FOR PER_ASG_ATTRIBUTE1 IS 'NOT ENTERED' DEFAULT FOR PER_ASG_JOB_ATTRIBUTE1 IS 'NOT ENTERED' DEFAULT FOR hr_assign_id is 0 Inputs are hr_assign_id l_hr_assignment_id = hr_assign_id CHANGE_CONTEXTS(HR_ASSIGNMENT_ID = l_hr_assignment_id) ( l_per_asg_location_name = PER_ASG_LOCATION_NAME l_per_asg_attribute1 = PER_ASG_ATTRIBUTE1 l_per_asg_job_attribute1 = PER_ASG_JOB_ATTRIBUTE1 ) RETURN l_per_asg_location_name, l_per_asg_attribute1,l_per_asg_job_attribute1 The next section of code is from the main calculation formula. It calls the formula, ODC_GET_HR_DATA above to retrieve the desired attributes: l_term_assignment_id = ASG_HR_ASG_ID SET_INPUT('HR_ASSIGN_ID',l_term_assignment_id) EXECUTE('ODC_GET_HR_DATA') Row = TRIM(UPPER(GET_OUTPUT('L_PER_ASG_LOCATION_NAME','NOT ENTERED'))) ASG_Shift = GET_OUTPUT('L_PER_ASG_ATTRIBUTE1','1') ASG_Job_Indicator = GET_OUTPUT('L_PER_ASG_JOB_ATTRIBUTE1','I') This example demonstrates the use of contexts to get data stored on the assignment screen. Other enhancements you will notice with elements and formulas are that you can now separate eligibility criteria from costing criteria. The final results of a calculation are stored on a separate indirect result element where costing can be attached. Eligibility can be attached to the main element. One other major change is that costing is performed at the same time as the gross to net calculation. There is no longer a separate Costing process to run. Wizards Many of the configuration tools are more wizard driven than completing a template screen. The Manage Element task is a good example. It steps you through the various options for creating elements. The wizard is smart enough to only ask the pertinent questions based on previous answers (i.e. you don’t have to complete any FSLA
- 4. COLLABORATE 15 Copyright @2015 by Hitachi Consulting 4 options when creating a deduction). All elements are created through this one task. Many of the transactional data is entered via wizards like new hire entry. Bank Account Management Fusion can now validate transit and routing number when entering personal payment methods. It is now a requirement to setup all banks and bank routing numbers before you can enter an employee’s bank account. We were able to use the Federal Reserve Bank Services website (https://www.frbservices.org/index.html), to download all of the Bank Names and Bank routing numbers. The Batch Loader task offers spreadsheet load capabilities to load the Bank Name and Bank Branch tables. Our naming standard for the Bank Branches was to make the leading characters of the Bank Branch be the routing number. For example, when loading bank information for Fifth Third Bank, we loaded the bank name (Fifth Third Bank) as the bank and 042000314-Fifth Third Bank as the Bank Branch name for the branch associated with routing number 042000314. This made the search process easy when selecting the correct branch for the desired routing number. One thing to watch here. When creating banks and bank branches, you cannot assign an effective date when loading this data. The system automatically assigns the session date when creating these attributes. This can cause problems when loading payment methods with effective dates before this session date. There is a payroll process parameter that can be set (Historic Payment Overrides data validation in payments for test environments) to override this validation. However, it only works in your test environments. Batch Loader One of the nice features of Fusion Payroll is the batch loader. It is similar to the BEE Spreadsheet loader in EBS, but with much more capabilities. As mentioned above, you can load attributes other than element entries with this tool, like banks, bank branches, assignment level costing and other useful batch loads. There are also Fast Formulas you can write that can parse flat files and load the resulting data in the batches. We used this functionality heavily to load Kronos time from the field into element entries. This was also helpful in processing inbound files like 401k deductions and certain benefit deductions from outside vendors. There is a useful document on My Oracle Support (Doc ID 1590004.1) that gives examples of the various load options. Payroll Flows One of the new concepts that users will need to grasp is the use of Payroll Flows. These flows are a series of tasks that are chained together in a single flow. A good example is the Quickpay flow. It includes the tasks calculating the quickpay, verifying results, running the prepayments, and then generating the resultant manual payment. Users can review each task before the next one starts. It provides a good checklist of items to perform to complete a series of tasks. There are delivered payroll flows for most of the main payroll processes, or you can run each task separately, similar to what we did in EBS. One concept that takes some time to get used to. Each iteration of a flow must be manually named and each name must be unique. We suggest using a naming standard to audit and validate your flows. Our standard followed this structure: YYYYMMDD XX <Payroll> Timestamp Where YYYMMDD was the date the process was run, or the payment date of the payroll process XX was an acronym that designated the process (QP for Quickpay, CP for Calculate Payroll, etc),
- 5. COLLABORATE 15 Copyright @2015 by Hitachi Consulting 5 <Payroll> was the payroll name associated with the process Timestamp was a date and timestamp to make the name unique. An example is the Calculate Payroll process for the Bi-Weekly payroll for the paydate of 7/4/2014: 20140704 CP Bi-Weekly 06-30-2014 9:44am Payroll Dashboard Fusion offers some interesting analytics via the Payroll Dashboard. It offers a complete picture of the status of various processes currently active. It includes pie charts that outline the various processes that are still in progress, detail flows that require attention and quick links to the details associated to the various pending payroll flows. We are still experimenting with these analytics to help us address outstanding issues that need to be resolved to complete an active payroll cycle. Challenges As with many new modules on the market, there are some areas where things don’t work as well as desired, or needed functionality is still lacking. This section will outline some of the challenges we uncovered while implementing Oracle Fusion Payroll. Data Loading There is not one consistent method of loading legacy data in Fusion. When operating in the SAAS model, you do not have access to the data APIs as you did with an on premise model. Therefore you cannot write your own PL/SQL procedures to load data. You must use the supplied tools to perform this important task. The two main tools at Release 8 were the File Base Loader and the Spreadsheet or Batch Loader. The File Base Loader uses a standard file layout for the corresponding data item. You must construct your legacy data to this template layout then execute the FBL process to load the data. The Batch Loader, as previously mentioned, resembles the BEE spreadsheet loader concept, where you copy and paste your data into a spreadsheet template. The table below outlines which tool we used for each data attribute: Entity Method Person, Address, Phone File Base Loader Work Relationship File Base Loader Payroll Relationship Spreadsheet Salary File Base Loader Contact File Base Loader Payment Method Spreadsheet Federal Tax Oracle Cloud State Tax Spreadsheet Element Entry Spreadsheet As noted in the table above, there are some attributes that you cannot load with any tool, such as the Federal Tax W4 elections and several descriptive flexfields. These must be loaded by Oracle Cloud Operations, using a template that they supply. Other items that Cloud Ops loaded for us included: US Taxation element (Vertex) Tax Reporting Unit
- 6. COLLABORATE 15 Copyright @2015 by Hitachi Consulting 6 Descriptive Flexfields on Person and Assignment Extra Information Types Local Taxes Overhead Door spent a considerable amount of time assigning the correct local tax for employees in Ohio and Pennsylvania. School Districts and PSD codes in Pennsylvania are loaded on the State W4 record, not on the local city or county as it was in EBS. We relied heavily on the website PA 32 to determine the correct PSD code and school district based on an employee’s work and residence address. This was a huge manual effort to touch every single PA and Ohio employee. In some cases we needed to assign a tax district on the employees main residence address in order for the proper tax to be deducted. We also encountered several scenarios where the proper tax district was not present in the pick list. This required several Service Requests to get Vertex and Fusion in line to properly tax employees. We also discovered that there is no validation in Employee Self Service when an employee enters a tax district on their home address. If an employee does key in an invalid entry, the next payroll process will error with an invalid geocode message. Also, if an employee changes any attribute on their address and they were currently deducting for a local school tax, someone in HR or Payroll must re-enter the correct school district back onto their record. Costing Overhead Door continues to use EBS for many of their ERP modules including Finance. One of our interfaces that we developed was sending the summarized costing data from Fusion to EBS. Standard Fusion functionality requires you to transfer the costing data into Subledger Accounting, write the corresponding accounting rules, then transfer the summarized journal entry into Fusion GL. Our hope was that we could retrieve this journal entry and then redirect it to EBS GL. However, we discovered that you cannot write the necessary accounting rules in SLA unless you had a valid Fusion GL license. Our only other option was to retrieve the costing data by running the Costing Detail report in Fusion Payroll and downloading it into a custom table in EBS. From there we could manipulate the data and create the necessary journal entry. The problem with this solution is that there is no interlocking capability without performing the Transfer to SLA. It becomes very easy to run the same costing report twice and hence posting the same costing results multiple times. We are trying to control this by selecting the proper dates, but the possibility of user error is great. We are working closely with Oracle Development to develop a much more controlled process to keep from duplicating journal entries. Final Thoughts Working in the Cloud can provide some interesting challenges as noted above. We highly recommend that if you are going to implement Fusion Payroll that you negotiate with Oracle during the sales cycle to get at least three pods or instances. Standard licenses only deliver a Production and a Test pod. We found that we needed more than that to control our UAT and Parallel test cycles. Also, you must plan early for your instance to instance copies. These must be performed by Oracle Cloud Operations. They require at least three weeks advanced notice before the copy can be scheduled. Even then, you may not get your desired date. We suggest that you work closely with your Center of Excellence coordinator if you cannot get your instances copied in time to meet your project plan milestones. Being a new module, we logged many Service Requests (SRs) to get critical issues resolved. It is important that you keep a close watch on these and escalate when you are nearing your critical path milestones. We held weekly SR meetings with our Center of Excellence coordinator to properly prioritize our open SRs so that we could get them resolved in time to meet our project milestones. We found that the sooner you can reach the actual development team, the quicker you receive the need workaround or solution.