Most businesses looking at HR audit services are looking at them for one of three reasons: a claim or an agency letter has arrived, a acquirer or investor is about to do diligence, or the business has grown past the point where one person can hold the whole picture. Those three want different things, and buying the wrong one is expensive in a way that only becomes obvious afterwards. A compliance-driven review, a diligence-driven review and a capability review differ in scope, in who they report to and in what the deliverable looks like, and the sensible first step is deciding which of the three you are actually buying.
The three reasons people buy, and what each should cost you
A compliance review is narrow and deep: specific record sets, tested hard, against the rules of your jurisdictions. A diligence review is broad and fast, aimed at anything an acquirer would raise, and its deliverable is a list of exposures with rough sizing. A capability review is about whether your HR function can carry the next stage of growth, and it is the only one of the three where the deliverable is largely opinion. Being explicit about which you want, in the engagement letter, is the difference between a report you act on and a report that sits in a folder.
What an outside reviewer genuinely brings
Independence, which matters when the finding is about the person who runs HR. Jurisdiction knowledge, which is the real value for any business employing people in more than one state or country, because the rules diverge in ways that are not intuitive. And pattern recognition from other businesses, which shortens the diagnosis: a reviewer who has seen the same classification problem twenty times identifies it from three files. None of that is available from a template, and all of it is wasted if the reviewer spends their first two days assembling your records.
The part to do before anyone is engaged
Assemble the population lists and the files. A current employee list from payroll, a leaver list for the retention period, and for each person the documents you expect to exist. Doing this yourself is not preparation in the cosmetic sense; it is the single largest cost item in the engagement, and it is billed at a rate you would not choose to pay for filing. It also produces most of the obvious findings for free, because the act of listing what should exist and ticking what does is exactly the test. The free audit preparation checklist on this site sizes that work in hours before you commit.
What to insist on in the deliverable
Findings tied to specific records rather than to general impressions, so the fix is identifiable. A severity or priority that is defined rather than felt. An owner and a realistic effort estimate for each remediation item, because a list of thirty findings with no sizing gets ignored. And a statement of what was and was not examined, including sample sizes: a report that does not say what it sampled cannot be relied on later, and cannot be compared to the next one.
Questions people ask about hr audit services
Is an HR audit confidential?
Only to the extent your engagement makes it so. Discuss with your own legal adviser how the report should be scoped and addressed before it is written, because that is much harder to arrange afterwards. This site does not give legal advice.
Can we just use a checklist instead of buying a service?
For record completeness, largely yes, and it is worth doing first. What a checklist cannot do is judge classification questions or multi-jurisdiction rules, which is where the exposure usually is.
How often should an HR audit happen?
Annually for the record completeness pass, which is cheap once the first one is done, and on any significant event: a new jurisdiction, a rapid hiring period, an acquisition, or a change in who runs payroll.
What should we send a prospective reviewer up front?
Headcount by location and worker type, the list of jurisdictions, and what triggered the review. That is enough for a realistic scope and quote, and it avoids the pattern where a fixed fee is quoted against a scope nobody had measured.