Every month it's the same ritual. Finance asks for labor cost by department. Payroll exports the report. Then somebody spends half a day in a spreadsheet, re-sorting rows, merging the two codes that both mean "Customer Service," splitting out the warehouse crew that's been sitting in "Operations - Misc" since 2019, and manually fixing the general ledger lines that never post where the controller expects. The report goes out late, with a caveat email attached.
Here's the uncomfortable truth: that's not a reporting problem. It's a structure problem. ADP Workforce Now will report on exactly what you've built, no better and no worse. If your departments, locations, and cost centers were set up in a hurry during implementation, or they've drifted through five years of reorgs, acquisitions, and "just add a code for now" decisions, then every report, every GL posting, and every headcount question inherits that mess.
The good news is that org structure is one of the highest-return cleanups in the entire system. Fix the foundation once and every downstream report gets faster, cleaner, and more trusted, permanently. Here's how to think about it.
Departments, locations, and cost centers aren't labels. They're the dimensions every dollar and every employee gets sliced by. When an employee is assigned to a department, that assignment flows into payroll registers, labor allocation, GL posting, headcount reports, benefit eligibility rules, security roles, manager dashboards, and often time and attendance approvals. One field, a dozen destinations.
That's why a sloppy structure can't be patched at the reporting layer. If two departments both contain sales reps because of an old reorg, no report writer setting will merge them cleanly forever; someone will hand-fix it every month until the structure itself is corrected. Conversely, a clean structure makes even simple, out-of-the-box reports immediately useful. The rule of thumb we give clients: if you routinely edit a payroll report in Excel before anyone can read it, the fix belongs in your structure, not your spreadsheet.
Structure also silently defines accountability. When cost centers map to real budget owners, managers can see and own their labor spend. When they don't, labor cost becomes "a payroll number" nobody feels responsible for, and budget conversations turn into archaeology.
These three levels get used interchangeably, and that's the root of many bad setups. They answer different questions.
Departments answer "what function does this person work in?" Think Sales, Engineering, Warehouse, Finance. Departments should mirror how the business actually organizes work and who reports to whom.
Locations answer "where does this person work?" Locations carry real compliance weight: they drive state and local tax jurisdictions, workers' compensation classifications, and state-specific policies. A location should be a real work site or, for remote employees, a deliberate convention you apply consistently. Never use locations as a stand-in for divisions or brands; the moment "location" stops meaning geography, your tax setup is at risk.
Cost centers answer "which bucket of money pays for this work?" They exist for finance, and they should map cleanly to your general ledger and budget structure. Sometimes cost centers align one-to-one with departments; often they don't, and that's fine, as long as the mapping is deliberate and documented.
The design test: any employee should be assignable with no debate. If assigning a new hire routinely triggers a discussion about which of two codes to use, the structure is ambiguous, and ambiguous structures decay fast.
The best structures are designed backwards from the reports people actually need. Before touching codes, list the ten questions leadership asks most: labor cost by department by month, headcount by location, overtime by supervisor, benefits cost by cost center, whatever they are. Your structure succeeds if those questions become one-click reports.
Then apply a few design principles. Keep each level to one meaning, function for departments, geography for locations, money for cost centers, and never encode two facts in one code. Use a consistent, human-readable numbering scheme with room to grow: leave gaps between codes so new departments slot in logically instead of getting the next free number. Match the depth of the hierarchy to the depth of your real reporting needs; if nobody will ever report on sub-teams within Accounts Payable, don't build codes for them. And write a one-page naming standard, because the structure you build this year will be maintained by someone else in three years.
Finally, decide who owns changes. Structure edits should have a gatekeeper, usually a partnership between HR and Finance, so codes are created deliberately rather than invented at 4:55 p.m. to get a new hire processed.
For most companies, the biggest payoff of clean structure is at the general ledger. Payroll is usually the largest expense on the P&L, and the GL interface in ADP maps earnings, deductions, taxes, and employer costs to accounts, sliced by your departments and cost centers. When the structure matches the chart of accounts logic, payroll posts automatically and the controller's month-end journal entry takes minutes.
When it doesn't match, you get the familiar pathology: payroll posts to suspense or default accounts, accountants reclassify labor cost by hand every month, and the GL mapping table grows special-case rules ("except for department 47, which splits 60/40...") until nobody dares touch it. Every one of those manual reclasses is a small monthly tax you pay for a structural decision made years ago.
Design the payroll structure and the GL mapping together, with Finance in the room. Confirm every department and cost center has a defined mapping, that new codes can't be created without one, and that employer taxes and benefits, not just wages, allocate the way Finance expects. If your accountants keep a "payroll reclass" journal entry template, that template is a map of exactly where your structure is broken. Read it.
Too flat. Everything lumped into six giant departments because that's what got set up at go-live. Reporting can only ever be as granular as the structure, so "Operations" with 140 people tells nobody anything. If managers keep asking payroll to "break that down further" and the answer requires a spreadsheet, the structure is too flat.
Too granular. The opposite failure: a code for every team, project, and two-person pod. Now every reorg requires mass employee transfers, reports need constant grouping to be readable, and half the codes have one or zero employees in them. Granularity feels like precision, but every level you add is a level someone must maintain forever. Projects and temporary initiatives usually belong in labor allocation or time tracking, not carved into the permanent org structure.
Stale codes. Departments that closed in 2021 still active in the system, three codes that all mean the same team because nobody merged them after the acquisition, terminated-location codes still attached to a handful of employees nobody rechecked. Stale codes don't just clutter dropdowns; they catch new hires. Someone picks the plausible-looking dead code, and that employee's cost lands in a bucket Finance stopped watching years ago.
A structure cleanup touches payroll, GL, security, and reporting at once, so treat it like a small project, not a settings change. The sequence matters.
Start with an inventory: every department, location, and cost-center code, its employee count, and when it was last used. Codes with zero employees and no activity in a year are your easy wins. Then design the target structure on paper with HR and Finance together, and get sign-off before anything changes in the system.
Build a crosswalk, old code to new code, for every employee and every GL mapping line. Time the cutover deliberately: the start of a quarter or the start of a year keeps before-and-after reporting clean, and you should avoid restructuring mid-quarter or during year-end processing. Update the GL mapping in the same change window as the employee transfers, not weeks later, or you'll have a gap where payroll posts against mappings that no longer exist.
After the first payroll under the new structure, reconcile hard: total labor cost by new department should tie back to expectations from the crosswalk, and the GL posting should land clean without suspense entries. Keep the old codes inactive but intact for historical reporting; deleting history is how you lose the ability to answer "how did this compare to last year?" And expect some year-over-year comparisons to need the crosswalk for the next four quarters; that's normal and worth the one-time cost.
A client with around 300 employees across four states came to us because their month-end labor reporting took their finance team three full days. The diagnosis was quick: 92 active department codes for what was functionally a 25-department company. Years of reorgs had left duplicate codes, dead codes, and departments that mixed function with location ("East Warehouse Admin"). Their GL posting used a mapping file with dozens of exception rules, and about $150,000 in monthly labor cost posted to a suspense account for manual reclassification.
We ran the cleanup over six weeks, timed to a quarter start. The inventory found 41 codes with zero active employees, retired immediately. The remaining structure was redesigned to 27 departments with one meaning each, locations restored to pure geography, and 18 cost centers mapped one-to-one against the chart of accounts. Every employee moved via a crosswalked mass update, and the GL mapping was rebuilt in the same window, exceptions eliminated.
The result: month-end labor reporting went from three days to about two hours, suspense-account postings went to zero, and, for the first time, department managers got a monthly labor report they could read without a translator. Nothing about the company changed. Only the structure did.
There's no magic number; there's a test. Every department should have a real manager who owns it, a reason Finance or HR would report on it separately, and enough employees to make its numbers meaningful. If a code fails all three, it probably shouldn't exist. Most mid-market companies land far closer to 25 than to 90.
You can restructure at any quarter start; January is convenient but not required. What you should avoid is mid-quarter cutovers, which split a single reporting period across two structures, and any restructuring during year-end processing, when the system and your team are busiest.
Department and cost-center changes are allocation changes; they move where cost is reported, not what anyone is paid. Location changes are the exception: because locations drive tax jurisdictions and workers' comp assignments, treat any location change with the same care as a new-hire tax setup, and verify withholding on the next check.
A healthy structure is one you rarely think about. New hires slot into an obvious code without debate. Payroll posts to the GL with no suspense account and no monthly reclass entry. Leadership's standard questions are standard reports. There's a one-page naming standard, a named gatekeeper for new codes, and a once-a-year review, an hour each January, that retires dead codes before they accumulate. When a reorg happens, the structure change is part of the reorg plan, executed at a quarter boundary with a crosswalk, instead of trailing it by a year.
If your structure has drifted and every report requires a spreadsheet rescue, this is one of the most satisfying projects we do. Ignite HCM consultants are former ADP service professionals serving ADP clients exclusively; we've designed and rebuilt these structures across hundreds of setups, and we know how to cut over without disturbing a single paycheck. No tickets, no hold queues — a dedicated consultant from inventory through first clean close.
Tired of fixing the same report every month? REQUEST A CONSULTATION.
ADP and the ADP logo are registered trademarks of ADP, Inc.
