Most ADP Workforce Now problems don’t come from software breaking. They come from a setup that made sense at go-live and slowly stopped matching how the company works. New states, new pay plans, new integrations and staff turnover all change what the system needs to do. The settings rarely change with them, and the gap shows up as extra work for the teams who run payroll.
This guide covers 6 signs that your ADP Workforce Now configuration has drifted far enough to need a formal review. Each sign comes with a way to measure it inside your own instance, so you can judge the problem with numbers instead of complaints. The later sections explain how to run the review and who should own the fixes once you find them.
TL;DR
• Rising manual checks and off-cycle runs usually point to a setup problem upstream.
• User access and security roles need regular review, because ADP’s controls assume you remove people who no longer need access.
• Integrations that need hand-fixing before every payroll have likely drifted from their original mapping.
• Earnings codes built at implementation may no longer match 2026 tax reporting or federal overtime rules.
• New tools such as ADP Assist, plus undocumented settings nobody can explain, are signs the instance has outgrown its last review.
What an ADP Workforce Now configuration review covers
An ADP Workforce Now configuration review is a structured check of how your instance is set up against how your business runs today. It looks at the settings your team controls. That includes pay rules, earnings codes, security roles, integrations, reports and workflows. It doesn’t audit ADP’s own infrastructure, since ADP maintains that under its own controls and audits.
The review usually starts where implementation left off. ADP’s published implementation overview describes a go-live process that includes business needs interviews, configuration, data conversion and a reporting and security review before launch. It also notes that ADP can convert up to 7 years of historical data. Those go-live decisions were made with the information available at the time. A review checks whether they still hold.
Scope depends on which modules you use. A company running only payroll will focus on pay codes and tax setup. A company using Time and Attendance plus third-party apps from the ADP Marketplace has more ground to cover. In both cases, the goal is the same. You want every setting to have a clear reason and an owner, and you want it to match current law and current operations.
A review keeps your current system in place. It finds the settings that cause rework or risk and fixes them where they are. Most of the 6 signs below can be confirmed with reports you already have access to.
Why these signs show up after go-live
The 6 signs share a cause: the business changed and the settings didn’t keep pace. New locations and bonus plans each add requirements, and they’re usually handled with a quick fix. The people who made those fixes move on, and tax and wage rules keep changing underneath them. Over time, the ADP Workforce Now configuration becomes a record of past decisions that no one can fully explain.
That’s why this guide focuses on symptoms you can observe and count. You don’t need to trace every past decision to know something’s wrong. If the signs below are rising, the setup has drifted, whatever the original reason was, and the numbers will show you where to look first.
Payroll corrections and off-cycle runs keep climbing
A steady rise in manual checks and off-cycle payrolls is the clearest sign your setup needs attention. Each correction means the system produced the wrong result the first time. When the same kinds of corrections repeat, the cause is often a setting in the ADP Workforce Now configuration. Late inputs, bad employee data and training gaps can also be at fault, so the next step is to sort one from the other.
Industry research shows how common this is. EY’s 2022 payroll error study surveyed payroll professionals at US companies with 250 to 10,000 employees. It found an average payroll accuracy rate of about 80%, and each error cost an average of $291 to fix. Time and attendance errors were the most frequent type, happening more than once per employee per year on average. Some errors cost far more, with direct deposit setup mistakes averaging around $1,500 per incident.
The survey also found that the average organization made about 15 corrections per payroll period. For a 1,000-employee company, fixing the 5 most time-consuming payroll errors added up to nearly 29 workweeks a year. Missing and incorrect time punches took the most time, at 26 minutes per employee.
How to measure it
You don’t need outside benchmarks to spot this sign in your own instance. ADP’s standard reports guide describes 2 reports that help. The Manual Check Details report lists every employee who received a manual check in the current pay period, along with the amounts paid. The Employee Payroll Changes report shows every change made to employee data in the period, with the old value next to the new one.
Time data deserves extra attention in this part of the review. The EY study found time and attendance mistakes were the most common payroll errors, and in ADP Workforce Now they often trace back to pay class rules or rounding settings. If the Employee Payroll Changes report shows the same timecard edits week after week, the pay class or the approval workflow likely needs a look.
Pull both reports for the last 6 to 12 months and group the corrections by cause. Look for repeats, such as the same pay code being fixed every cycle or the same department needing manual checks. Repeated payroll errors with a shared cause point to a setting. Random, one-off errors point more toward training or data entry. The first group is what a configuration review can fix.
Security roles and user access haven’t been reviewed
If nobody can say when user access was last reviewed, that’s a sign on its own. Access tends to grow over time. People change jobs and keep their old permissions. Contractors finish projects and never get removed. Admin rights get handed out to solve a one-time problem and never taken back.
ADP Workforce Now uses a layered access model. Predefined security roles, such as security master and security administrator, sit alongside security groups and custom profiles that control what each user can see and do. The security master is the most trusted role, with complete access to the system. That flexibility lets you fit access to each job. It also means access can spread in ways that are hard to see without a structured review.
Why this matters for audits
ADP publishes a SOC 1 report on its own controls, and many auditors rely on it. But a SOC 1 report assumes the client runs certain controls of its own, known as complementary user entity controls. An explainer on user entity controls gives the most common example. The service provider has no way to know when a client’s employee leaves, so the client must remove that person’s access. If the client doesn’t, a former employee can keep access to the system.
ADP runs this kind of control for its own staff. The company’s global security overview says it formally reviews its employees’ access rights at least once a year to confirm they still match each person’s role. Your users are your responsibility, and the same discipline applies.
Single sign-on adds another layer to check. One published ADP client agreement describes federated single sign-on, where the client’s own identity system controls provisioning and deprovisioning for ADP access. If your company uses it, access to your ADP Workforce Now configuration depends on your identity provider being kept current. A review should confirm that people removed from the corporate directory also lose access to ADP, and that any accounts created outside single sign-on are accounted for.
A public example shows what a good review looks like. A 2018 municipal audit report from the City of Markham examined its Workforce Now setup. The auditor noted that users get access through predefined security roles, and recommended reviewing all roles for segregation-of-duties conflicts. The city responded by starting a quarterly review of users with special access. That’s a useful model. Pull a list of every user and their assigned roles, confirm each person still works for you and still needs that access, and flag anyone who can both change pay data and approve payroll.
Integrations need manual cleanup before every payroll
If your team fixes imported hours or rates by hand before every payroll, your integrations have probably drifted. A healthy integration should need checking, but it shouldn’t need regular repair. Routine cleanup is a sign the mapping between systems no longer matches how either system is set up.
The problem often comes down to which system is responsible for what. Help documentation for When I Work’s ADP integration shows how this plays out. The tool sends position-specific pay rates into ADP as a Temporary Rate. It doesn’t change rates for overtime premiums, which have to be set up in ADP Workforce Now itself. If both teams assume the other system handles overtime, nobody does.
Other tools have similar limits. ClockShark’s documentation says that when an employee tracks time to a task linked to a pay rate that isn’t assigned to them in ADP, the system uses their base rate instead. The hours arrive correctly, so the problem is easy to miss. The rate behind those hours is wrong, and it stays wrong until someone checks.
What to check
Start by listing every system that sends data into ADP or pulls data out of it. For each one, write down what data moves and which codes or fields it maps to. Then compare that list with what’s actually configured on both sides. Pay special attention to anything added or changed since go-live, such as new positions or new pay rates.
Next, track how much manual work each integration causes. Ask the payroll team to log every fix they make to imported data for 2 or 3 pay cycles. The log will show which integrations cause the most rework. Those are the ones to fix first. This part of an ADP Workforce Now configuration review often saves the most time, since the same fix pays off every payroll.
Earnings codes and pay rules no longer match the law
Earnings codes built at implementation don’t update themselves when wage and tax rules change. If your code list hasn’t been reviewed in the last year, there’s a good chance some codes no longer do what the law requires. This sign carries the most legal risk of the 6, because the errors affect pay and tax reporting directly.
The regular rate problem
Federal overtime law requires overtime at 1.5 times the regular rate of pay. The regular rate often includes more than the base wage. The DOL’s Fact Sheet 56C explains that nondiscretionary bonuses must be included in the regular rate unless another exclusion applies. That covers many attendance and production bonuses.
Whether a bonus feeds the overtime calculation depends on how its earnings code is set up. A code created years ago may not include it. The cost of getting this wrong can be large. In May 2025, the DOL announced that Cupertino Electric paid more than $1.4 million in back wages to over 2,600 employees for leaving nondiscretionary bonuses out of the overtime rate. Across all cases, the DOL’s Wage and Hour data shows more than $259 million in back wages recovered in fiscal year 2025.
The 2026 qualified overtime change
A newer issue affects W-2 reporting. The One Big Beautiful Bill Act created a federal tax deduction for qualified overtime for tax years 2025 through 2028. Only the premium portion of overtime required under the FLSA qualifies. An employer guide on the overtime deduction notes that overtime required only by state law isn’t qualified, so no part of it is deductible for federal purposes.
This changes how earnings codes need to be set up. ADP’s 2026 tax season update says that for check dates of January 1, 2026 or later, hours in the existing overtime and double time codes are used to calculate the qualified amount for Box 12 of the W-2. ADP added new “NQ” (non-qualified) codes for overtime that doesn’t qualify. A new Qualified Overtime Compensation Summary report shows the qualified premium per payroll.
State-only overtime needs its own treatment. California daily overtime, for example, isn’t qualified when it’s owed only under state law. Some of those hours can still count as FLSA overtime if the employee also passes 40 hours in the workweek. If you run all daily overtime through the standard codes, ADP may treat hours as qualified that aren’t, and your W-2 figures can come out too high. ADP’s earnings quick reference lists the NQ codes and notes that those amounts don’t qualify under the new law. A review should map every overtime scenario you pay to the right code.
How to review the code list
Export every active earnings code and deduction code. For each earnings code, confirm what it’s for, whether it’s still used and whether it should feed the regular rate. Flag codes with no activity in the past year as candidates for retirement. Look for duplicates, such as 3 different codes that all pay the same kind of bonus. A clean code list makes every other part of payroll easier to check.
New features like ADP Assist were never turned on
A review should also look at what the system can do that your team isn’t using. SaaS platforms change every year, and features added after go-live often go unused because nobody on the team knows they exist. If your payroll checks look the same as they did 5 years ago, you may be doing by hand what the system can now flag for you.
The best current example is ADP Assist. In September 2025, ADP announced new AI features for Workforce Now built on ADP Assist. One of them is anomaly detection, which identifies inconsistencies in payroll data and suggests corrections for practitioners to review and approve. ADP says early adopters report saving up to 30 minutes per payroll cycle. That figure comes from ADP, so treat it as a vendor claim until you measure it yourself.
The same announcement covered conversational reporting. Users can ask a plain question, such as how overtime costs are trending, and get charts and answers back. For teams that wait days for custom reports, that can change how quickly problems get noticed.
Adding new features carefully
A new feature is only useful if it fits your setup. Anomaly detection works best when the underlying data is clean. If your code list is cluttered or your integrations send bad data, ADP Assist will flag a lot of noise along with real problems. That’s why it makes sense to fix the earlier signs first, then turn on new tools.
Ask your ADP account team which features are included in your current package and which cost extra. Then test each one on a single pay group before rolling it out. Measure the result against your baseline correction count, so you know whether the tool actually reduces payroll errors in your environment.
Nobody can explain why settings exist
The last sign is the hardest to measure and often the most telling. If your current team can’t explain why a pay class or pay code exists, the setup has outgrown its documentation. That makes every other problem slower to fix, because nobody knows what will break if a setting changes.
This usually shows up in small ways. A new payroll manager finds 4 bonus codes and doesn’t know which one to use. An HR admin copies an old pay class for a new location because it seems to work. An IT lead can’t find the person who set up a benefits carrier feed. Each case adds risk, because the next change is made without full knowledge of the last one.
Where to start
Begin with ADP’s own configuration reports. The standard reports guide lists a Time and Attendance setup report that shows items such as holidays, pay cycles, pay classes and security groups in one place. Use it as a starting inventory. Then go through each item with the people who use it most and write down what it does and why it exists.
Anything nobody can explain goes on a separate list. Some items will turn out to be needed. Others can be retired after a test. The result is a living record of your ADP Workforce Now configuration that the next person can read. That record is often the most lasting benefit of the whole review.
How to measure each sign in your own instance
Each of the 6 signs can be checked with data you already have. The table below maps each sign to a starting metric and where to find it. Use these as a baseline before the review and measure again afterward.
| Sign | What to Measure | Where to Find It |
| Payroll corrections | Manual checks and off-cycle runs per period | Manual Check Details report |
| Access sprawl | Users with admin roles and inactive accounts | User and security role listings |
| Integration cleanup | Manual fixes to imported data per payroll | Payroll team log over 2 or 3 cycles |
| Outdated codes | Active codes with no use in 12 months | Earnings and deduction code export |
| Unused features | Features included but not turned on | ADP account team and release notes |
| Missing documentation | Settings with no known owner or purpose | Time and Attendance setup report |
Collect each number the same way every time. If one person pulls the manual check count from a report and another counts it by hand, the trend won’t mean much. Write down the exact report and date range you used, so the comparison after the review is fair.
The numbers don’t need to be perfect. Their job is to show whether each problem is getting better or worse. A baseline also helps when you ask for budget or time to fix something, since you can show the cost in hours and corrections.
When you present the baseline, connect it to cost. Using EY’s average of $291 per error as a rough guide, a team that makes 15 corrections per biweekly payroll would log 390 corrections a year. At $291 each, that’s about $113,000 in correction costs before any fines or back pay. Your own numbers will differ, but the math helps leaders see payroll errors as a budget issue.
Some teams also track the time payroll takes from start to approval. If that number grows while headcount stays flat, it usually means more manual checking is happening. That’s a useful early warning, even before corrections start rising.
Reading the results fairly
Raw counts can mislead. A company that grew from 400 to 600 employees will see more manual checks even if its setup improved. Divide each count by headcount, such as corrections per 100 employees per payroll, so growth doesn’t hide progress or make things look worse than they are.
Watch for seasonal spikes too. Year-end processing and annual bonus runs add corrections that have nothing to do with drift. Compare the same months year over year, or leave those periods out of the baseline so they don’t skew the trend.
Look at the mix as well as the total. A drop in total corrections is good news, but if the share tied to a single pay code keeps growing, that code still needs attention. Grouping corrections by cause, as described in the payroll corrections section, makes the mix easy to see and easy to explain to leadership.
Turning the signs into a review plan
Once you have baseline numbers, let them decide where the work starts. The signs with the worst or fastest-rising numbers set the first scope. That keeps your ADP Workforce Now configuration review tied to evidence you can measure again afterward.
1. Score each sign from your baseline and note whether it’s rising or falling over the last 6 to 12 months.
2. Pick the 1 or 2 signs with the clearest problem as your first scope.
3. For each sign in scope, trace a sample of cases back to a cause, such as a specific pay code or integration field.
4. Separate setting causes from process or training causes, since only the first group is fixed by changing the ADP Workforce Now configuration.
5. Fix the setting and test the change on a small group before rollout.
6. Measure the same sign again after 2 or 3 payroll cycles, using the same report and date range as the baseline.
7. Move to the next sign once the first one’s numbers improve.
Working one sign at a time keeps changes to live payroll small. It also makes cause and effect easy to see. If corrections drop after an earnings code fix, you know the fix worked, because nothing else changed in that window.
Deciding what to fix first
Not every finding carries the same weight. Rank each one by 2 questions: could it cause wrong pay or wrong tax reporting, and how often does it create manual work? Items that touch pay accuracy or legal compliance go first, even if they rarely come up. Access problems come next, because a single wrong permission can expose pay data for the whole company. Items that only cost time, such as a slow report or an extra approval step, can wait for a later cycle.
A simple 2-column list works well here. Put each finding under pay and compliance risk or under time cost, and note its owner beside it. This keeps the ADP Workforce Now configuration review focused on the problems that matter most. It also gives leadership a clear picture of what was fixed and what’s still open.
Working with ADP
Bring in ADP early for anything you can’t change on your own. Some settings, especially in tax and payroll setup, may need ADP’s support team to update them. Knowing which changes need ADP helps you plan timelines and avoid surprises.
Who owns the fixes after the review
A review only helps if the fixes stick. That depends on clear ownership after the project ends. In a SaaS model, ADP runs the platform and the client owns the setup. Your team decides which rules apply to which employees and who has access.
Split ownership by area. Payroll should own earnings codes and pay rules. HR should own employee data standards and security roles, with sign-off from someone outside payroll for sensitive access. IT should own integrations and single sign-on. Name a single person to coordinate across the groups, so changes don’t slip through the gaps between teams.
Change control keeps new drift from building up. Without it, the ADP Workforce Now configuration you just cleaned up will slowly pick up the same kinds of quick fixes that caused the problems in the first place. The review becomes a one-time event, and the next team starts over from scratch. Treat any change to pay rules or integrations like a software release. Write down what’s changing and why. Test it on sample employees before it goes live. Keep a record of who approved it, so the next review starts with a clear history instead of guesswork.
Finally, set a schedule. A light check each quarter on access and corrections catches problems early. A deeper review once a year, timed before year-end tax work, keeps pay codes and reporting current. The payroll errors and access issues that build up between reviews are much easier to fix when they’re caught within a few months.
Final thoughts
An ADP Workforce Now instance rarely fails all at once. It drifts, one quick fix at a time, until payroll takes longer and nobody is sure why settings exist. The 6 signs in this guide give you a way to spot that drift early and measure it with data you already have.
The good news is that most fixes happen inside your existing setup. Cleaning up pay codes and repairing integrations can cut rework without a new system or a large project. A regular ADP Workforce Now configuration review, backed by clear owners and simple change control, keeps the system matched to how your business actually runs.
Frequently asked questions
What is an ADP Workforce Now configuration review?
It’s a structured check of how your instance is set up against current business needs and current law. It covers the settings your team controls, such as pay rules, security roles, integrations and reports.
How often should we review our ADP Workforce Now configuration?
A light quarterly check on user access and correction volume works well for most companies. A deeper annual review, done before year-end tax work, keeps pay codes and reporting current.
What are the most common signs a review is needed?
The most common signs are rising manual checks and off-cycle runs, unreviewed user access, integrations that need hand-fixing, outdated pay codes, unused new features and settings nobody can explain.
How much do payroll errors cost?
EY’s 2022 study found an average cost of $291 per payroll error, with an average payroll accuracy rate of about 80%. Some errors, such as direct deposit setup mistakes, averaged far more per incident.
Which ADP reports help find payroll problems?
The Manual Check Details report shows who received manual checks in a pay period. The Employee Payroll Changes report lists changes to employee data with old and new values. Together they show where corrections are happening.
Does ADP review our user access for us?
No. ADP secures its platform and reviews its own staff access, but your company controls its own users. Removing former employees and reviewing access are client responsibilities.
What are complementary user entity controls?
They’re controls a service provider expects its clients to run so the provider’s own controls work. A common example is removing system access when an employee leaves, since the provider can’t know about the departure.
Why do integrations cause payroll errors?
Integrations can send data with the wrong rate or map hours to the wrong code. Some tools also leave parts of the calculation, such as overtime premiums, to be set up in ADP. If neither side handles it, the result is wrong.
How do earnings codes affect overtime?
Each earnings code decides whether a payment feeds the regular rate used for overtime. If a nondiscretionary bonus code doesn’t feed it, overtime will be underpaid every time that bonus is paid.
What are the NQ codes in ADP?
NQ means non-qualified. ADP added these codes for overtime hours that don’t qualify for the federal overtime deduction, such as overtime required only by state law. Hours in standard overtime codes are treated as qualified for W-2 reporting.
What is ADP Assist?
ADP Assist is ADP’s AI-enhanced platform. In Workforce Now, it includes anomaly detection that flags inconsistencies in payroll data and suggests corrections, plus conversational reporting for workforce questions.
Should we turn on ADP Assist before cleaning up our setup?
It’s usually better to clean up first. Anomaly detection works best with clean data, and a cluttered setup can produce many flags that aren’t real problems.
How long does a configuration review take?
It depends on how many modules and integrations are in scope and how much documentation already exists. Setting a firm scope at the start keeps the project from growing.
Who should be on the review team?
Include at least 1 person from payroll and 1 from HR. Add IT if you run integrations or single sign-on. Name a single owner to coordinate decisions and keep the project on schedule.
Do we need ADP’s help to make changes?
Some changes, especially in tax and payroll setup, may need ADP’s support team. Many others, such as user access and documentation, your team can handle directly. Confirm which is which early in the project.
