TL;DR
- Core decision: Whether to publish a single global band set at the strictest rule, or genuinely different postings by location.
- When doing nothing is right: None of the reader's hiring locations trigger a pay-range rule, and the reader's ATS isn't configured to surface ranges anyway.
- What has to be true: Whichever option is picked, the published band must be the band the hiring manager would actually defend in a pay conversation.
- How the options split: Single global band favours operational simplicity and consistent candidate experience; differentiated postings honour each rule but force dual content operations.
- Decision rule: If the reader hires in five US states and the EU, the marginal cost of running two content operations is usually less than the legal and trust cost of publishing one band wide enough to satisfy all of them.
- Expected outcome: Either path can work, but only if the band posted is a band the company would actually pay within, and only if existing staff are warned before the postings go live.
The First Posting the Board Sees
A reward lead is staring at a single job description for a senior product manager. The role is open to candidates in Chicago, Minneapolis, Boston and Berlin. The careers site template only has one salary field. It's the first posting the new CHRO has asked to see before the board meeting on Thursday, and the reward lead has just been told that what gets entered here will be the company-wide template from now on.
She drafts a band she thinks is defensible, then a wider band in case a candidate in any of those markets pushes back on it, then a still wider band because Vermont's rule sits at five employees and she can't quite remember whether the rule is the band or the description of benefits. She closes the laptop at midnight with three drafts and no decision. The range in the widest one is broad enough that a candidate in any market could be inside it, and that's exactly why no candidate in any market should trust it.
But the real issue isn't which number to type into the field. It's whether the company wants to run a single global pay policy or a federation of location-specific ones, and whether the careers site, the ATS and the hiring manager training are set up to do either. Picking a number without that decision first is the move that fails.
Best tools for Payroll & Compensation
When You Do Not Actually Need to Act Yet
There's a real version of this reader who can sit this out. Four shapes of it.
The single-location team. A twenty-person product company hiring only from one US city, no remote postings, no plans to change. None of the listed state triggers apply because none of those states are involved. The careers site doesn't need a salary field at all. The reader's actual job in this case is to keep it that way: if the company stays single-location, no disclosure rule fires. If it plans to hire in a listed state next quarter, the work starts before the first posting is published. The posture is genuinely low-effort, but it's not no-effort, because the moment the company posts a remote role it has not thought through, the posture is gone.
The multi-state team with no listed state yet. A reader hiring across Texas, Florida and Washington, with no current postings in Illinois, Minnesota, New Jersey, Vermont or Massachusetts. Today, no disclosure rule in this set applies. The trap is treating that as permanent. Several of these laws expanded coverage to remote applicants in ways that are easy to miss, and the company that hires its first Vermont candidate tomorrow without a band is the company that learns the rule the hard way. For now, the practical work is mapping which state the next hire might be in, and watching that map. A reader who treats the gap as a free pass has to keep checking, because the moment a single applicant sits in a covered state the calculation flips.
The team whose ATS makes the wrong default. A reader whose ATS shows a salary field, prefilled with a generic band pulled from a generic band. If that field is visible to candidates and the band doesn't match a good-faith expectation, the company is publishing a range that's hard to defend. The fix in this shape isn't a new policy. It's turning the field off until the policy exists. The temptation is to leave the field visible with a placeholder and revisit it later, but the placeholder is itself the publication. The field is either off, or it's a real band that the company stands behind.
The EU team waiting on transposition. A reader in a member state that has not transposed yet. The Directive's deadline has passed and only four states met it. Public-sector employers in late-transposing states face vertical direct effect, meaning the Directive applies to them directly where national law is silent. Private employers are in a different position: national courts must read existing national law in a way that's consistent with the Directive, where that's possible. That's not the same exposure as a fully transposed regime, and it's also not zero. A reader in this shape has time to prepare, but not time to ignore. The preparation that pays off now is mapping the Directive's floor (history ban, secrecy clauses unenforceable, joint pay evaluation above a five per cent gap) and checking which pieces the national system already gives them, because the work needed at transposition is much smaller if those pieces are already in place.
The Five Questions You Are Asking Yourself at 11pm
Do I really have to publish a range in every posting? Only if a posting is for a role that can be performed by a candidate in a listed jurisdiction, by a reader who employs enough people to be covered, and at a point in time after the relevant effective date. If any link in that chain is missing, the rule doesn't fire. The answer for a given posting turns on location, headcount and date, not on the company's preference. The reasoning behind it is that each of these statutes is written to a specific fact pattern, and the legal trigger is the fact pattern, not the policy intent. If the reader gets this wrong by assuming any of the three conditions is satisfied when it isn't, they publish a band they didn't have to and signal to candidates and competitors something they didn't mean to signal. The flip side is the reader who assumes none of the three is satisfied and skips a posting that did need a band, and that mistake is the one a regulator notices first.
Can I publish one band wide enough to satisfy everyone? Mechanically, yes. Practically, no. A band built to cover every rule and every market is too wide to be a good-faith range in any of them. Good faith is read by reference to what the employer expects to pay, not to what the employer is willing to type into a field. A band that contains every plausible outcome for every market is, by definition, not what the employer expects to pay for any single hire. If the answer to "what would you actually offer this person?" is anywhere inside the band, the band has failed its only test. The reasoning is that good faith is comparative: it asks what a reasonable employer in that market would pay for the role. A band that runs from entry to senior is not what a reasonable employer would pay for a single hire, and a regulator reading the posting will see that. If the reader gets this wrong, the band gets challenged, the company's postings get re-read line by line by counsel, and every other posting the company has live gets questioned in the same review.
Is the EU one rule or twenty-three? Right now, twenty-three. Four member states have transposed, and the rest haven't. The Directive sets the floor: pay history questions are out, pay secrecy clauses are unenforceable, joint pay evaluations follow a gender pay gap above five per cent. Each member state writes its own implementing law on top of that floor. A reader hiring in two member states is reading two implementations. A reader hiring in five is reading five. The reasoning is that a directive is a result the member state has to reach, not a text the member state copies, which means every implementation is a translation, and translations drift. If the reader treats the EU as one rule, the company runs the risk of publishing a band that's compliant in one state and non-compliant in the next, and the cost of sorting it out after the fact is the cost of doing the work twice.
Will my existing employees compare their pay to the new postings? Yes, immediately, and usually inside a week. This is the consequence teams consistently fail to plan for, and it's the one that does the lasting damage. A published range that sits below what a current employee earns is a public statement that the company underpays them. A range that sits above is a public statement that the company would replace them. Neither is comfortable, and both have to be managed before the postings go live. The reasoning is that pay is one of the few numbers employees have never been able to compare, and the moment the comparison is possible the comparison happens. If the reader gets this wrong and posts without warning, the first internal conversation is happening in a corridor, the second is happening with HR, and the third is happening with a lawyer, and none of those conversations recover the trust that was lost in the corridor.
Can I keep asking about salary expectations? In the listed jurisdictions, asking candidates about pay history is restricted or banned, and several of the US rules in this set also restrict expectation questions. In the EU, the ban is on history, not expectation. The shape is the same across the set: the employer's range is meant to lead the conversation, not the candidate's last number. If the application flow asks for current salary or recent comp, it has to change. The reasoning is that the regimes share an underlying logic: the candidate's previous pay should not anchor the new one, because previous pay carries bias forward. If the reader keeps the question in the flow, the question itself is the breach, regardless of what the company does with the answer, and the breach sits in a database the regulator can subpoena.
Three Honest Categories the Approaches Split Into
Single global band, set at the strictest rule. The reader picks the strictest rule that applies to any of their postings and writes one band that satisfies it everywhere. Operationally clean. One field, one content operation, one training programme for hiring managers. The risk is that the strictest rule's requirements bleed into postings where they don't apply, in ways that look strange to candidates and to local counsel. A Berlin candidate seeing a US-shaped description of benefits, or a Vermont candidate seeing a band that was built for a different market, is a small paper cut that adds up. The single global band works when the reader's hiring footprint is genuinely narrow, or when the strictest rule is so close to the others that the differences don't show. A concrete example is a 200-person company hiring in one US state and one EU member state, where the same band happens to clear both rules because the markets are close enough that the bands are close enough. In that case the operational simplicity is real and the paper cuts don't appear. The category fails when the markets diverge enough that the band has to grow to cover both, and the band stops being a band and starts being a placeholder. It also fails when local counsel flags that a posting in their jurisdiction should not contain language from another, and the global template has to be forked anyway.
Differentiated postings, one rule per location. The reader maintains genuinely different postings for each listed jurisdiction, with bands and language that match the local rule. This is the only path that honours what each statute actually says. The operational cost is real: two content operations, two ATS configurations, two sets of hiring-manager scripts, and a versioning problem the moment a rule changes. It works when the reader is large enough to absorb the overhead, when local counsel can review the postings, and when the company has decided that candidate trust is worth the cost of running it twice. A concrete example is a 700-person company hiring in Illinois, Minnesota, New Jersey, Vermont, Massachusetts and three EU member states, where the reward team owns a posting template per jurisdiction, the ATS supports per-location overrides on the same requisition, and hiring managers have a jurisdiction-specific script for pay conversations. The cost is the second content operation, and the benefit is that each posting reads as written for that market. The category fails when the company lacks the governance to keep the postings in sync as rules change, because a stale posting is worse than a single global band that was honest about its limits. It also fails when the ATS cannot support per-location overrides and the workaround is two separate requisitions for what is functionally the same role, which doubles the candidate's experience of the company.
Withhold range entirely, where the rule allows. In a small set of circumstances, mostly where the reader doesn't employ enough people in a given listed jurisdiction or the role genuinely can't be performed there, no range has to be published at all. The temptation is to stretch this category further than it goes. "We aren't sure we have to" isn't the same as "we've confirmed we don't." A reader who treats a gap in coverage as a permanent exemption is the reader who gets caught when the next remote hire is in a listed state. The category is real and narrow. Outside it, the only choices are the first two. A concrete example is a 12-person team in a US state that isn't covered, hiring only locally, with no remote pipeline: the careers site has no salary field, the application flow doesn't ask for compensation, and the only pay conversation happens at offer. That posture is correct today. The category fails the moment the company posts a remote-friendly role, or hires its first employee in a listed state, or gets approached by a candidate who sits in a covered state. The transition out of this category has to be designed in advance, because the first posting under the new rule will arrive faster than the policy.
Five Diagnostic Questions to Self-Assess Against
Where, exactly, are the next twenty hires going to sit? Pull a real list. Don't answer in the abstract. The list, sorted by state or member state, is the rule-set the reader is actually facing. Any answer that starts with "we hire across the US and Europe" has already failed. The way to actually answer this for the reader's own organisation is to take the open requisitions, tag each one by intended location, and add the roles the team expects to open in the next two quarters. Where the list is uncertain, mark it uncertain, because uncertain is its own bucket. The list drives everything downstream: which jurisdictions fire, which template the ATS needs, which adviser reviews which posting.
What is the smallest band the hiring manager for each role would actually defend in person? Have the hiring manager say the number out loud, for a specific candidate, for a specific level. The band published is the band the manager will say. If the manager wouldn't say the published number, the band is wrong. To answer this for the reader's organisation, sit with each hiring manager, give them a candidate profile and a level, and ask what they would offer. The number they say is the midpoint of the band. The band is the midpoint plus or minus a defensible step. Any band wider than that has to be justified in writing, and "we wanted flexibility" is not a justification that survives contact with counsel.
How does the ATS handle a single role with multiple locations? Open it. Try to set up a role that's open in two states with two different bands. If the system can't do it without two separate postings, the reader's architecture has answered the policy question for them. To actually answer this, log in, create a test requisition, and try to attach two bands to it. The path through the system either supports it or doesn't, and the answer determines whether differentiated postings are a policy choice or an architecture fact. If the ATS can't do it, the reader is either living with single global bands or running two requisitions, and the second option has consequences for the candidate experience.
What is the existing internal pay range for the same role, by location? Pull it. If the published external band is wider than the internal one, the company is telling candidates something it doesn't believe. If it's narrower, the company is promising something it can't deliver. The two ranges have to be consistent with each other, not just with the rule. To answer this for the reader's organisation, ask the reward team for the internal range document, sort it by role and location, and lay it next to the proposed external band. The gap, if there is one, is the conversation the reader has to have with the hiring manager before the posting goes live.
Who is going to tell existing employees what changed, and when? Name the person. Pick the date. If the answer is "we'll figure that out after the postings go live", the consequence has already been chosen, and it isn't a good one. To answer this for the reader's organisation, the answer has to be a name with a date and a draft of what will be said. If the name is missing, the date is missing, or the draft is missing, the answer is "no", and the reader has to fix one of the three before the postings go live. The internal communication is not a side artefact; it is the artefact that determines whether the launch is an event the company manages or an event that happens to the company.
The Jurisdictions, Compared
Illinois
Illinois requires employers with fifteen or more employees to disclose a salary range and a general description of benefits in postings, with effect from January 2025. The rule applies at a low headcount threshold relative to its peers, which makes it the trigger that fires earliest for many growing companies. Its real bite, beyond the band itself, is the requirement to describe benefits. A band with no benefits language is non-compliant even when the band is correct, and a generic benefits paragraph lifted from another state's posting will read as boilerplate. The reason this entry earns a place at the top is that Illinois is where most readers find out they have a posting problem: it triggers at a headcount that catches companies before they've hired a reward team, and it forces a benefit description that other states don't.
Where it genuinely falls short is in handling remote roles. The rule's text speaks to postings with a connection to the state, and the connection for a remote-first role isn't always obvious. A reader hiring remotely into Illinois from another state has to decide whether the connection is the candidate, the role, or the company's footprint, and the answer varies by counsel. The statute doesn't give a clean remote-posting test, so the reader is making a judgement call their adviser will have to defend.
Minnesota
Minnesota applies at thirty or more employees and took effect in January 2025. The threshold sits higher than Illinois's, which means some readers who are covered in Illinois aren't covered in Minnesota and the two postings look different for a reason that's hard to explain to a hiring manager. It earns its place because it is the rule readers most often miscopy from Illinois, on the assumption that a similar rule is the same rule. It isn't, and the assumption shows up in postings that look compliant until a regulator reads them.
The rule's weakness is the same as Illinois's, sharpened: good faith is the operative phrase, and what good faith means for a band isn't defined numerically. A reader who posts a band that's wide enough to cover any candidate they might hire has not posted a band; they have posted a placeholder. Regulators reading the rule have a lot of room to call that out, and candidates reading it have a lot of room to distrust it.
New Jersey
New Jersey took effect in June 2025. The rule's distinctive feature, relative to the others in this set, is the way it handles internal postings and promotions. A reader whose first instinct is to treat this as a candidate-facing rule has to widen the lens to internal movement as well, because the rule speaks to postings broadly. It earns its place because internal postings are the rule readers forget about until HR runs the first internal move under the new regime and discovers the band has to be there too.
Where it falls short is in specifying what counts as a posting for a role that may be performed in New Jersey but is filled remotely from elsewhere. Again, the remote-posting test is the gap. A reader with a single New Jersey employee but a national remote workforce is in a fact pattern the rule doesn't address cleanly, and the reader has to choose between conservative and liberal readings and accept the cost of whichever they pick.
Vermont
Vermont applies at five or more employees and took effect in July 2025, and it requires disclosure before or at the time an offer is made. The five-employee threshold is the lowest in this set, which puts it inside the reach of very small employers, and the timing rule sits at the offer, not the posting, which is the rule's distinctive feature. It earns its place because Vermont is the rule that breaks the "we're too small to worry about this" assumption, and the timing rule forces an offer-stage process the company has to be ready to run.
The weakness is the timing. A disclosure that arrives at offer has to be supported by a process the company can run reliably at the offer moment, including for candidates who were screened under a different disclosure regime earlier in the funnel. A reader who treats the Vermont rule as a separate item of paperwork at offer is the reader whose offer process falls apart at scale, because the disclosure has to match what the candidate saw earlier in the process or the candidate notices.
Massachusetts
Massachusetts applies at twenty-five or more employees and took effect in October 2025. It sits between Illinois and Minnesota in threshold, and its requirements are broadly similar to its neighbours, which makes it the rule a reader is most likely to assume they have already handled because they handled a nearby one. It earns its place because Massachusetts is the rule that punishes copy-paste: similar enough to look handled, different enough to fail.
The weakness is the assumption. Massachusetts has its own posting language, its own good-faith standard, and its own interaction with internal postings. A reader who copy-pastes an Illinois posting into Massachusetts has done the work that looks like compliance without doing the work that's compliance. Each jurisdiction reads its own rule against its own postings, and the differences are small until they're not.
The EU Pay Transparency Directive
The Directive's transposition deadline of 7 June 2026 has passed. Only four of the twenty-seven member states met it: Slovakia, Italy, Lithuania and Malta. The Commission has said there will be no extension, no pause and no carve-out. The Directive sets a floor across the union: a ban on asking candidates about pay history, an end to pay secrecy clauses, a joint pay evaluation where a gender pay gap above five per cent is found, and gender pay gap reporting on a cycle that begins with data collection now and first reports due in June 2027. It earns its place in any comparison because no US state rule looks like it, and a reader hiring in the EU is operating under a regime the rest of this article doesn't prepare them for.
Its weakness, for a multi-state reader, is that it isn't one rule. It's a directive implemented twenty-three different ways. The Directive tells each member state what the result must be. Each member state writes the route. A reader hiring in two member states is reading two implementations, and the differences show up in the bands, in the reporting, in the timing of disclosure, and in the definition of worker representative.
The Decision Table
| Situation | Scale | Setup | Primary Pain | Recommended Starting Point |
|---|---|---|---|---|
| Hiring only in one US city, no remote | Under 15 employees everywhere | No salary field needed | Future-proofing if footprint grows | Document the single-location posture and set a watch date for the first listed state |
| Hiring in two US listed states, no EU | 50 to 250 employees | ATS supports one field per posting | Two content tracks and two hiring-manager scripts | Differentiated postings, one rule per state, with a single benefits paragraph where both rules allow |
| Hiring in three US listed states and the EU | 250 to 1,000 employees | Multi-region ATS, multiple legal advisers | Running a federation well | Differentiated postings by jurisdiction, a single internal pay-range policy that underpins all of them |
| Hiring in five or more US listed states and the EU | Over 1,000 employees | Global TA, in-house reward | Maintaining two parallel content operations at scale | Differentiated postings, plus a governance forum that owns cross-jurisdiction consistency |
| Hiring remotely from anywhere, footprint unknown | Under 30 employees | ATS allows any-candidate applications | Determining which jurisdictions fire | Conservative posture: post a band on every external posting until footprint is locked |
| Hiring in an EU member state that has not transposed | Any size | Local counsel, no national implementing law yet | Uncertainty between vertical direct effect and directive-consistent interpretation | Map the Directive's floor and apply it where national law is silent; prepare for transposition |
| Mostly single-jurisdiction but plans remote hiring next quarter | 20 to 50 employees | ATS configured for one location | Locking in a template before the second location arrives | Hold the template; design the ATS to support per-location overrides before they are needed |
| Internal pay ranges disagree with the proposed postings | Any size | Existing salary bands by location | Trust damage when staff compare | Reconcile internal ranges first, then publish; never publish a band that contradicts what staff already have |
Setting a Range You Can Defend
A good-faith range is one that genuinely reflects what the employer expects to pay. The phrase is short, and the work behind it's long. The employer has to be able to say, for a specific candidate at a specific level, that the offer would fall inside the published band, and that the band was not constructed after the candidate's level was known.
The starting point is the internal salary range for the role. If the company has one, it has the spine of the band. The external range is a publishable version of the internal one, adjusted only where market data makes the internal range genuinely indefensible to candidates. Where the internal range is honest, the published range should be close to it, not a wider shape built to give hiring managers room to move. Room to move is what an internal range is for. A published range that's wider than the internal one is a range the company doesn't believe.
A defensible band has three properties. It's narrow enough that a candidate could reasonably ask why they have been placed where they have. It's wide enough that two candidates at the same level could be paid differently for reasons the company can articulate. And it overlaps with what the company's existing staff at the same level are paid, so that the act of publishing doesn't create an immediate internal incident.
An indefensible band has three properties too. It's wider than the internal range, which means it contains outcomes the company wouldn't actually offer. It's built around the rule rather than around the role, which means the band serves the posting rather than the hire. And it contradicts what existing staff are paid, which means the company is publishing a public statement about its own people that those people are about to read.
| Element of the band | Defensible | Indefensible |
|---|---|---|
| Width | Two or three steps, market-tested | Wider than the internal range for the same role |
| Anchor | The internal salary range for the role | The union of every market the company hires in |
| Treatment of bonus and equity | Either inside the band with a stated basis, or outside with a clear note | Inside the band with no basis, or outside with no note |
| Existing staff | Overlapping with what current staff at the same level earn | Materially below or above what current staff earn |
| Updates | Revised when the role, the level or the market shifts materially | Set once and carried forward through several hiring cycles |
| Geographic treatment | A band tied to a specific market the role is actually performed in | A band written to satisfy every market the role might touch |
| History of revisions | Each revision documented with a rationale in the file | Revisions made without a traceable reason |
What Changes Internally the Day You Publish
The external posting is the visible half of the decision. The internal half is what happens inside the company on the day the postings go live. An existing employee at the same level as the posted role will, sooner or later, find the posting. They will compare their own pay to the band. They will form an opinion about whether they're paid fairly. That opinion will be more or less fixed by the time they speak to their manager, and the manager will be more or less prepared for the conversation.
The first thing that breaks is the assumption that pay is private. It isn't, once it's published, and the team has to behave as if every employee has seen every band for their level. That means the internal ranges, the explanation for any differences, and the path to address gaps have to exist before the postings go live. They can't be assembled in response to the first internal complaint.
The second thing that breaks is the manager's confidence. Managers who were trained to deflect pay questions now have a published range to point to, and the deflection skills they were taught before publication are the wrong skills for the conversation after. A reader who has not retrained the manager cohort before the postings go live is going to lose those conversations, and the manager is going to learn in front of the employee.
The third thing that breaks is the promotion conversation. An employee at level three looking at a level four posting they think they could do is going to ask what the band is for the level they should be at next. If the band is published, that conversation is happening now, with a number attached. If it isn't, the employee is going to ask when it will be.
The fix for all three is the same: publish internal ranges alongside external ones, train managers on the new conversations, and time the external launch so that the internal conversation has already happened in private.
| Internal consequence | What it looks like | What has to be in place first |
|---|---|---|
| Staff comparing their pay to the band | Quiet conversations, then questions to managers | Internal ranges consistent with the published ones |
| Manager pay conversations retraining | Existing scripts stop working | New manager training, delivered before launch |
| Promotion expectations reset | Staff ask about bands at the next level up | Career-level bands published internally |
| Pay equity questions | Staff compare across demographic lines they had not compared before | Pay equity analysis run before the postings go live |
| New-hire offers questioned by existing staff | Current staff ask why new hires are paid inside a band they sit outside | Offer rationale documented, and an internal escalation path for gaps |
| Recruiting referrals drying up | Existing staff stop referring candidates because they don't want to be compared to the band | Referral policy reviewed, and a clear statement on how internal pay relates to external ranges |
| Attrition risk in bands that read low | Staff in lower-paid cohorts start updating CVs | Retention review for roles where the band sits below local market |
What to Put in Writing
A good decision without records is a decision that can't be defended later. The artefacts below are the ones a regulator, a candidate, or a plaintiff's counsel will eventually ask for. They're also the ones an internal audit will ask for when the policy is reviewed.
The single rule behind the list is that every decision about a published band should leave a trace. Not because every trace will be read, but because the traces that get read are the ones the reader didn't know would be read. The board approval, the range rationale, the manager sign-off, the candidate-facing version, the internal announcement, and the retention schedule each protect a different failure mode.
The retention question matters more than it looks. Pay records are kept for years in most regimes, and the document that explains why a band was set the way it was is the document that defends the band years later. The reader who treats the rationale as a one-off conversation in a chat thread is the reader who can't reconstruct it when it matters.
| Artefact | Who owns it | When it is written | What it prevents |
|---|---|---|---|
| Board or executive approval of the policy | Reward lead, with CHRO sign-off | Before the first posting goes live | Accusations of unilateral policy change |
| Range rationale, per role, per location | Reward team, with hiring manager input | When the band is set and when it is revised | "The band was arbitrary" allegations |
| Hiring-manager sign-off on each posting | Hiring manager, attested in writing | Before the posting is published | Managers disclaiming the band they posted |
| Candidate-facing posting, version controlled | TA operations | At publication and at every revision | Drift between approved and published versions |
| Internal communication, sent before external launch | Internal comms, with reward input | One to two weeks before launch | Staff discovering the change from candidates |
| Pay equity analysis | Reward, reviewed by legal | Before launch and at each annual cycle | External publication surfacing an internal gap |
| Retention schedule for all of the above | People operations | At creation of each artefact | Inability to produce the artefact when asked |
| Incident log of internal questions and answers | People operations | Live, from launch onwards | Repeated questions landing without an institutional memory |
| Adviser sign-off, per jurisdiction, per template | Reward, with external counsel | At template creation and at each rule change | Postings published without a lawyer having read them |
Questions to Ask Before You Commit
Coverage. Which jurisdictions are we actually hiring in over the next twelve months, and which of those trigger a disclosure rule? A bad answer is a list of cities instead of states, or a list of states without a date.
Bands. For each role, what is the smallest band the hiring manager would defend in a pay conversation with a specific named candidate? A bad answer is a band wider than the internal range, or a band the manager can't quote without checking.
ATS. Can our system support different bands by location for the same role, and can it version the change? A bad answer is "we'd have to do that manually" for a system the reader expects to use at scale.
Internal pay. How does each published band compare to what current staff at the same level are paid? A bad answer is "we'll check after launch".
Manager readiness. When did our hiring managers last practise a pay conversation using a published band, and what did they get wrong? A bad answer is "we'll send a script".
Staff communication. What will existing employees be told, by whom, and when? A bad answer is "the day the postings go live".
History and expectation questions. What does our application flow ask candidates today, and which of those questions are we now restricted from asking? A bad answer is silence, or "we'll review the form".
Reporting and evaluation. For EU postings, what is our reporting calendar, and who owns the gender pay gap report when it's due? A bad answer is "we've time".
Governance. Who owns the policy on paper, who owns it in the ATS, and who owns it when the rule changes? A bad answer is three different people.
Adviser. Which adviser has reviewed a draft posting from each jurisdiction in the last quarter? A bad answer is a name without a date.
The Cost of Getting This Wrong
The visible cost is the one the legal team will quote. A posted band that doesn't match a good-faith expectation is a band a regulator can challenge. The exposure is real, and it's the cost that makes the project urgent.
The cost that never appears on an invoice is the cost to candidate trust. A candidate who reads a band that's too wide will discount the whole posting, including the role and the company. A candidate who reads a band that's too narrow will assume the company isn't serious about flexibility. Either way, the candidate's read of the company is shaped by the band before the hiring manager has said a word.
The cost that compounds is the cost inside the company. An employee who finds the band and decides they're underpaid is an employee who has just formed a view that the company is treating them unfairly. That view doesn't go away when the band is revised. It goes away, if at all, when the company acts on it visibly and quickly. So the question the reader has to sit with isn't "what band do we post" but "what does our company look like to the people already inside it when that band goes live".
When You Are Ready to Go Further
If the reader has reached this point, they have a map of the problem. The next move is to pressure-test it. HROpsLab's independent comparison work covers the platforms and advisers that operate across these jurisdictions, and the editorial team publishes what each one is genuinely good at and where each one falls short. The point of the comparison is to make the reader's shortlist shorter, with evidence behind every cut.
For readers who want a second opinion on a specific configuration, the editorial team runs office-hours conversations that aren't sales calls and don't become ones. The point is to hear the reader's situation and tell them what the comparison work suggests for it. HROpsLab is a review publication, not a vendor, and the team sells nothing.
Frequently Asked Questions
Can a single posting satisfy every state with one band?
Mechanically, yes, but the band has to be a good-faith range in each jurisdiction it covers, and good faith is read against what the employer expects to pay in that market. A band wide enough to satisfy every rule is rarely a band the employer would defend for any single hire. Differentiated postings are usually the cleaner answer, and the only path that survives a regulator reading the posting against the role rather than against the rule.
What does "good-faith range" actually mean?
It means a range that genuinely reflects what the employer expects to pay for the role, not a placeholder written to satisfy the rule. The hiring manager has to be able to defend the band in a pay conversation with a specific candidate. If the manager wouldn't, the band isn't in good faith, and the gap between the published band and the manager's actual offer is the gap a regulator or a plaintiff's counsel will use to argue the band was a sham.
Should the band include bonus and equity?
Either include them with a stated basis for how the number was reached, or exclude them with a clear note. Including them with no basis, or excluding them with no note, both undermine the band's credibility. The reader's choice between the two depends on whether the variable component is set by formula or by judgement, and the published band has to match whichever process actually runs at offer time.
What do we do about remote roles open to any state?
Pick the conservative posture: post a band on every external posting until the footprint is locked. Then map where the next twenty hires will actually sit, and let that map drive which jurisdictions the ATS has to support per role. The remote-posting test isn't clean in any of these statutes, so the safe move is to assume the rule applies, because the cost of over-posting a band is much smaller than the cost of under-posting one.
Will existing staff ask about the new postings?
Yes, immediately, and usually inside a week. The act of publishing a band turns pay into something staff can compare to their own. The reader has to communicate internally before the postings go live, with internal ranges that are consistent with the external ones, and the internal conversation has to happen in private before the public one happens in the open.
Can we still ask about salary expectations?
In the listed US jurisdictions, expectation questions are restricted or banned depending on the rule. In the EU, pay history is restricted, not expectation, but the spirit of the regime is the same: the employer's band leads the conversation, not the candidate's last number. The application flow has to change before the next posting goes live, because the breach sits in the data the regulator can subpoena, not in what the company does with the answer.
What if the band we post turns out to be wrong?
The band has to be revisable, with a documented rationale for each revision. A band that's set once and carried forward through several hiring cycles is a band that drifts away from what the company actually pays. The reader's governance has to include a regular review, not a one-off posting, and the review has to be on a calendar the company can defend when the review didn't happen.
Do we need different postings for different EU member states?
Right now, yes. The Directive sets a floor across the union, but each member state writes its own implementing law, and four have transposed. A reader hiring in two member states is reading two implementations, and the differences show up in bands, in timing, in reporting and in the definition of worker representative, which means a single EU template is not a compliant template at scale.
HROpsLab gives reward leaders the evidence to make defensible decisions, without selling them the tools to do it.
