TL;DR: Talent acquisition is a specialist discipline within HR — but that doesn't settle where it should sit on your org chart. This guide sets out four realistic reporting models and the conditions at 50–500 headcount under which each one works.
The short answer — and why it's the wrong question
Yes. Talent acquisition is a specialist discipline inside the people profession, not a separate function by definition. The CIPD Profession Map lists resourcing — identifying, attracting and recruiting people — as specialist knowledge alongside reward, employee relations and talent management. That is the definitional answer, and if it is what you came for, you can stop reading.
Here's the problem. Knowing that TA is "part of HR" tells you nothing about who your recruiters report to, who holds the hiring budget, or who a frustrated hiring manager escalates to when a req has been open for eleven weeks. Those are the decisions an HR Director actually faces, and the definitional answer does not touch any of them.
The stakes are higher when hiring is hard, and in the UK it currently is. The CIPD/Omni RMS Resourcing and Talent Planning Report 2024, as summarised by Talk Staff, found that 84% of UK organisations tried to fill a vacancy in the past year and 64% had difficulty attracting suitable candidates.
So this article treats the question as what it is: an organisation-design choice. Four models, the conditions that favour each, the obligations that stay fixed whichever you pick, and a checklist you can run against your own data. Scope is deliberately narrow — UK employers between 50 and 500 people, spanning the official medium band (50–249) and the lower part of the large band, per GOV.UK business population estimates. That is where the answer genuinely changes.
TA and HR: same profession, different operating rhythm
Plain definitions first, because the search results are full of muddled ones.
Talent acquisition (TA) covers attracting, sourcing, assessing and hiring — employer brand, pipeline, selection, offer, candidate experience. Generalist HR covers the full employment lifecycle from contract to exit: terms and conditions, absence, performance, grievances, exits. Talent management (TM) is broader than both. The CIPD's talent management factsheet describes it as the systematic attraction, identification, development, engagement, retention and deployment of people who are valuable to an organisation — which makes acquisition the first stage of talent management rather than a rival to it.
It helps to see where this sits in the shape of an HR function generally. UK employers at this size usually run one of three: a single generalist team covering the whole lifecycle, a business-partner model where partners sit close to the business and specialists sit centrally, or — as headcount climbs — centres of excellence for reward, resourcing and employee relations behind a shared-services layer. TA is one specialism within that, so the question of where it reports is a slice of the broader structural question rather than a separate one, and the shape of the rest of your function narrows the sensible answers before you start. For the wider picture, we've mapped the seven main functions of HR separately.
The two disciplines feel different day to day, and that is not imagined. TA runs on external signals: pipeline velocity this week, what a competitor is paying, whether a market has gone cold. Generalist HR runs on internal case load, policy consistency and precedent — the answer you give one employee has to survive being quoted back at you by the next one. Both rhythms are legitimate. The friction between them is real and it is structural, not personal.
That friction is where the argument for independence comes from. The search firm Elliott Scott HR makes the case that TA sitting inside HR drifts towards a transactional mindset — filling chairs rather than building pipelines. It is a fair observation, and worth taking seriously. It is also a practitioner argument rather than a benchmarked finding, made by a firm that sells executive search. That does not make it wrong; it does mean you should weigh it as advocacy.
The counterweight sits in the same CIPD/Omni data: 41% of UK organisations saw a new hire resign within 12 weeks. Some of that is bad sourcing. A good deal of it is a handover problem — the point where recruitment ends and induction begins, and nobody quite owns the gap. Separating TA from onboarding has a cost, and that figure is roughly what it looks like.
Four ways UK companies actually structure this
Most coverage frames this as binary: TA inside HR, or TA independent. In practice it is a spectrum with four recognisable points.
Model 1 — Generalist-owned. No dedicated recruiter. The HR Manager or People Lead hires alongside everything else, with agencies used for the roles that are genuinely hard. Typical at 50–120 headcount with under roughly 20 hires a year. Cheap, well-coordinated, and it falls over the moment hiring volume spikes, because recruitment is the thing that gets dropped when an employee relations case lands.
Model 2 — Specialist inside HR. One or more in-house recruiters reporting to the Head of People, drawing on the HR budget, working in the same ATS and HRIS stack as everyone else. This is the most common shape in the 150–400 band. You get recruiting expertise without losing the continuity from offer through to probation, and compliance sits under one leader.
Model 3 — Dotted-line hybrid. The TA Lead has a solid line to the business — a COO, a CRO, a divisional director — and a dotted line to HR for process, compliance and systems. Almost nobody writes about this model, which is odd, because it is frequently the right answer at this size. It buys stakeholder proximity without severing the compliance and data connection. It also demands the most maturity: dotted lines that are assumed rather than written down quietly become no line at all.
Model 4 — Independent function. A Head of TA with their own budget and their own headcount plan, reporting to the CEO or COO. Justifiable when hiring volume is the primary constraint on growth and hiring performance is a board-level topic. Worth noting that CEO sponsorship is a precondition here, not a nice-to-have — the CIPD/Omni summary found 46% of UK organisations said their CEO prioritised talent management, and without that sponsorship an independent TA function tends to get orphaned — no budget champion when the next cost review lands.
Budget ownership, not the box on the org chart, is what actually changes behaviour.
That last point deserves emphasis. A recruiter who cannot commit spend without an HR sign-off behaves like Model 2 regardless of where the line is drawn. Conversely, a recruiter sitting formally inside HR with a ring-fenced sourcing budget and direct hiring-manager accountability behaves a great deal like Model 3. If you are trying to change behaviour, follow the money before you redraw the chart.
Should talent acquisition report to HR?
Models 1 and 2 tend to win under a recognisable set of conditions: hiring volume under roughly 25–30 hires a year; replacement hiring dominating growth hiring; a lean team where one more specialist head is a genuine budget trade-off; a single site or a single country.
The retention argument is the strongest one. If your problem is early attrition rather than pipeline volume, keeping sourcing, onboarding and the first 90 days under one accountable leader is worth more than recruiter autonomy. You cannot fix a handover by giving one side of it a different boss.
Then compliance. At 50–250 headcount you rarely have a second person properly trained on Equality Act obligations in the hiring process. Concentrating that expertise in HR is a real risk control, not bureaucracy — and it is the control most often lost when TA moves out.
Cost matters too. A dedicated TA function carries fixed cost through hiring troughs. Where demand is lumpy, TA inside HR flexes down through agency spend rather than through a redundancy conversation.
For instance, take a 180-person UK professional services firm hiring 20 people a year across two offices, mostly backfilling qualified staff. Splitting TA out would give you one recruiter with a business reporting line, no meaningful volume to build pipelines against, and a new coordination seam between hiring and induction. You would buy coordination cost and no speed.
When should TA sit outside HR?
Models 3 and 4 come into their own under different conditions: hiring is the rate limiter on revenue; sustained volume above roughly 40–50 hires a year; a small number of business-critical role families where speed and market intelligence beat process uniformity; multi-country hiring where local market knowledge outweighs central policy.
The strongest genuine case is stakeholder proximity. Recruiters embedded with the function they hire for get better calibration, faster interview feedback and earlier warning when a role spec is about to change. That is hard to replicate from a central HR team serving eight departments.
The budget-protection argument cuts both ways. An independent budget insulates hiring investment from HR cost pressure — useful when the board is squeezing overheads and the pipeline is the thing that suffers. It also removes a check on cost-per-hire, and agency spend is exactly the line item that drifts without one.
Now the failure mode nobody writes about. TA moved to the business without a deliberate dotted line back to HR loses three things quickly: access to and governance of the ATS and HRIS data, the separation of equal-opportunities monitoring from decision-making, and the offer-approval controls that keep pay bands intact. None of these break loudly. You find out at the tribunal, or when you discover two people in the same band are materially apart on pay. The split has to be designed, not declared.
There is an adjacent governance point worth flagging. TA teams are among the fastest adopters of AI sourcing and screening tools, and an independent TA function buying its own tooling is precisely the scenario where an unreviewed screening algorithm enters the business without HR or IT seeing it. If that is your direction, read up on how UK recruiters are using AI across the hiring funnel and decide who reviews the tools before someone buys one.
What doesn't change, whoever TA reports to
No UK law prescribes where TA sits. ACAS is explicit that there are no recruitment processes required by law. This is a pure organisation-design choice, and nobody is going to tell you off for it.
What does not move is that the statutory obligations attach to the hiring activity, not to the department performing it. Equality Act 2010 duties run through adverts, shortlisting, interview and offer, including the duty to make reasonable adjustments for disabled candidates at any stage of the process. ACAS sets out the practical detail.
Two traps catch a business-reporting TA function more often than an HR-embedded one. The first is asking about health or disability before an offer is made, which is generally not permitted. The second is failing to keep equal-opportunities monitoring data separate from application materials and out of the hands of the people making the decision — an easy control to lose when the ATS configuration moves to a team with different priorities.
Candidate data is regulated employment data under UK GDPR. Sourcing lists, screening notes and rejected-candidate records are all in scope, and the retention clock does not pause because the recruiter now reports to the COO.
The practical instruction is simple. Whatever model you choose, name one person accountable for recruitment compliance and write it into their objectives. In Models 3 and 4 that person usually still sits in HR — so make the dotted line explicit rather than assumed.
A decision checklist for HR Directors at 50–500
Answer these six with your own data before you restructure anything.
- Annual hire volume, and its trend over the last three years — not this quarter.
- Growth hires versus replacement hires, expressed as a ratio.
- Time-to-hire on your three most business-critical role families, not the company average, which hides everything interesting.
- Twelve-week attrition rate for new starters.
- How many countries you hire in, and whether local market knowledge is the binding constraint.
- Who currently signs off recruitment spend — including agency fees.
Be clear about what those thresholds are worth. There is no UK dataset that maps headcount to TA reporting structure, so the numbers in this article are reasoned starting points drawn from operating conditions, not benchmarks. Anyone who tells you to split TA out at exactly 500 employees is repeating folklore, most of it imported from US tech.
On sequencing: change budget ownership and hiring-manager accountability before you change the org chart. Most of the benefit attributed to splitting TA out comes from those two things, and both are reversible in a way a restructure is not.
One caution. Do not restructure in response to a single bad hiring year. The CIPD/Omni summary found 69% of UK organisations reporting increased competition for skilled candidates, most acutely for senior and skilled roles. A hard year is the market, not necessarily your structure. If you want to look at the functions around TA before making the call, we've published a self-assessment checklist for your core HR processes.
The takeaway
Talent acquisition is part of HR as a discipline — the CIPD's own Profession Map settles that. Where it reports is a separate decision, driven by hiring volume, the mix of growth and replacement hiring, and how directly your revenue depends on filling specific roles. It is also reversible, which should make it less fraught than it usually is.
The four models are a spectrum rather than a binary, and the dotted-line hybrid is the most under-used of them. At 50–500 headcount it is frequently the answer that gets you business proximity without losing compliance and data governance.
Whichever you choose, the Equality Act and UK GDPR obligations stay with you, so name who owns them in the new shape.
Before you move anything, run the six checklist questions against your own hiring data — most HR Directors find the answer is already sitting in numbers they have.