Farming out Payroll Fundamentals
Back in the good ol' days of payroll you skipped out of the field on the 15th and last day of each moon cycle and someone handed you three shekels. So endeth the pay-period. Fast forward a few thousand years and you are now looking over the pdf of your paystub wondering what the heck paycode ‘22FL150 - 2/3rds 150% Funeral Leave Pay' is (not to mention, who died?).
Way back when, it was instantly apparent if you were paid correctly or not. Now, you aren't so sure. Don't worry the people paying you can't be sure either. Feel better? I didn't think so. While I can't tell you the specific screw up on your paycheck or those within your company I would like to cover some fundamentals people forget while toiling away in some silo of the HR/Payroll plantation. Ignoring these fundamentals, in my opinion, has lead to overly complex time & pay tracking mechanisms that make payroll and/or their associated accounting errors (typically 1%-8%) very easy to create and very difficult to find.
Fundamental #1: It is COMPANY data- not Payroll Data, or HR Data or Time & Attendance data.
The second you start thinking attendance, timesheet, job cost or other information can be divorced from its other genomes in the helix you are going to end up with two-headed sheep! (Okay-farm analogy ends here). Do not let someone run wild in some aspect of the system (timecards, paycodes, GL Interface and the like) thinking they are isolated from all the others (not my job syndrome... don't you just love this pic!) 
Don't let HR escape the meeting you are having on state taxes for the new sales employees in Montana. "That's a finance issue" is the wrong answer. And make sure HR brings the latest pay policy manual to the meeting-we will need it for Fundamental #2.
Fundamental #2: Payroll policies don't generate paychecks-Data Entry People and Lines of Code do.
If your HR VP ever says "The payroll company is handling it-I gave them our payroll policy manual and they are taking care of everything" WHACK him or her on the head with the payroll policy manual. (then ask if you can split the dollar value of each error you find*). I once watched as my Kronos techie did the math for the HR VP. While I agree my techie shouldn't have used the phrasing "Any fifth grader can see THAT policy on THIS system will result in a 13% OVERPAYMENT", my techie's math was irrefutable.
The HR VP refuted it and said something like "I think (big name payroll company) is capable of handling our simple little company's payroll" and shuffled off with his policy manual. The lawsuits threatened when (big name payroll company) tried to get the overpayments back after more than 6 months of errors were impressive. At least those errors were consistent and replicatable. Want some variety? Take 2 or 3 or 10 payroll entry people, put them in different rooms and give them the same timecard/overtime/vacation pay situation and see how many different and creative ways it gets coded. Should be just one, right? It won't be for all but the simplest of entries. Do this regularly.
Fundamental #3: Different levels of granularity are needed by different departments at different time intervals with differing precision.
People just don't get this and they often come up with really bad schemes in order to implement really dumb pay policies that may not even be mathematically possible. In its simplest form, this means that while PayChexR-Us need only know the employee worked 8 hours to pay correctly, the system may need to split that into two different departments, or paycodes, or some other distribution mechanism to do some other calculation correctly. People maintaining these mechanisms often use the wrong tool for the job or, more commonly, don't understand the ramifications elsewhere in the system for changes that appear to get the result they wanted. More complex examples of this principle have to do with HOW percentages are calculated, WHAT rounding must be done, WHERE totals are needed, and WHEN periods are defined. This is closely tied to Fundamental #1 and either results from asinine pay policies or in asinine pay algorithms. You can't always control the former but you had better manage the latter.
Fundamental #4:
Making any changes to the HR/Payroll machinery without being able to model and test for effect across all stakeholders is Baaaaaaaaaad! (Hey-I didn't say no sheep allusions). Holding to this fundamental discipline is simply the last chance to check all silos before dollars go into bank accounts, tax authorities get paid, re-imbursement invoices get generated, etc. I'm not saying generate an entire mock payroll for every little change but people who really KNOW how the change will affect their area (check every area) need to be consulted and an overall model reviewed every time. Yes, every time. Walk everyone thru it from the CFO down to the shop floor supervisor with his hands on the time clock. (The shop floor guy, I have found, is more likely to identify real problems with the latest change).
That is it. Just four but I'm constantly amazed out how many people involved in the paying for and tracking of people attendance are not grounded in these fundamentals. Of course, straying from them is what keeps Kronos, ADP, labor attorneys and others in this business IN business. No sense in paying us a shekel more than necessary.
About the author: Jeff Millard has over 25 years of IT professional experience, he has a proven vision and leadership ability beyond mere technical expertise. He is well versed in a broad range of best-practice approaches, he has a unique ability to chart clear courses for growth oriented companies.