Workplace massage: the offer, tiers and pricing
The argument for on-site massage as part of a workplace mental-health stack is laid out in detail at the case. This page is the offer side: what a session looks like, what it costs, how it sits in a corporation tax return and a PAYE Settlement Agreement, and what scale of commitment matches what scale of need.
Four rungs are available, from a half-day on-site to a sustained programme. They differ in cost, in scale, and, critically, in the tax route they sit inside. The first two rungs use the trivial benefits exemption at ITEPA 2003 s.323A, which works only at small per-head amounts and only for genuinely one-off bookings. The upper two rungs use a PAYE Settlement Agreement, which lets the firm absorb the income tax and Class 1B NIC on the employee’s behalf. The choice is not stylistic; the tax architecture sets it.
The offer
Same therapist every visit. Self-book on a shared calendar. No HR mediation, no disclosure required, no note on anyone’s file. Invoice in arrears, on a per-session basis, so the cost flexes with take-up. Either side steps out with two weeks’ notice. Heated table, fresh linens, oils, candles, music; setup in ten minutes, packdown in five. £5m public liability cover, commercial motor insurance, DBS cleared, GDPR compliant.
The four tiers
Two paths into the offer. A one-off (top row of cards) introduces the therapist and the format to a team or floor; under the trivial-benefits route at small per-head amounts it sits outside the employee’s payslip entirely. A committed engagement (bottom row) builds a deliberative trail under the HSE’s Working Minds framework and sits inside a PAYE Settlement Agreement so the firm absorbs the tax.
What the pilot looks like
Not every employee wants a massage. The ones who do will come forward; the rest are working alongside someone whose week has changed. Most firms run this as a twelve-week trial: a therapist on-site weekly or fortnightly, six to ten one-hour slots per visit, employees self-book on a shared calendar. No HR mediation, no disclosure required, no note on anyone’s file. Invoice in arrears, on a per-session basis, so the cost flexes with take-up. Either side can step out with two weeks’ notice. If the numbers don’t move, the trial ends. If they do, the conversation becomes about how often, and for how many.
At £69 per session attended, the twelve-week pilot at eight slots per visit caps at £6,600 in direct cost. The PSA gross-up addressed below adds roughly 43%, putting the fully-booked ceiling at approximately £9,500. These figures assume recipients in the £25–60k threshold band, the workforce this is designed to reach, not partners, who are typically positioned to fund private remedies themselves. In practice, a fresh pilot does not arrive at the ceiling. A reasonable forecast based on cohort data is take-up between 30% and 50% over the twelve weeks, putting the all-in cost in the region of £3,000 to £4,800.2 Below the threshold at which most procurement functions engage. A single resignation in this salary band costs the firm approximately £43,200 in today’s money: recruitment, training, and lost output during the 28-week ramp-up to optimum productivity combined.1 One resignation prevented covers the pilot nine to fourteen times over.
Take-up is materially affected by how the offer is communicated. A clear booking system, transparent session lengths, and consistent slot timings produce engagement that ad-hoc taster events do not. The pilot’s framing matters as much as its content.
The cost is the small part. The firm thinks the person is worth spending money on as a person, not as a unit of output.
What firms typically do
A firm wanting to test before committing books a Half-day for a single team or floor. The format under the trivial-benefits route is structurally one-off, so the question of “does this work for our people” gets answered without any tax-architecture or PSA-application work up front. If the answer is no, the conversation closes. If the answer is yes, the question becomes how often and for how many, and the discussion moves to Pilot.
A firm acting on a Working Minds prompt, or on its own foreseeability assessment, tends to go straight to Pilot. The deliberative-quality test the HSE applied to Birmingham is a documentation test, not an outcome test: a policy, an anonymised booking log, and a quarterly aggregate uptake report at modest scale read as a defensible position. A Half-day, however well-received, does not produce that documentary trail.
Pilots that produce visible uptake convert to Sustained at the firm’s own pace. Pilots that do not, end at twelve weeks with no contractual tail. Either result is useful information.
What documentation looks like
For Tier 3 and Tier 4 engagements, the firm receives a per-visit booking log (anonymised by default; full identification available if the firm prefers and the employees consent), a quarterly aggregate uptake report covering session counts, slot fill rate, and the cumulative roster of unique recipients, and a written provision policy in template form for the firm to adapt and adopt formally. Tier 4 adds an annual review document benchmarking the year’s uptake against the firm’s policy targets.
These artefacts are sized to support a Working Minds reasonable-steps evidence portfolio. They are not a substitute for the firm’s own stress risk assessment, workload monitoring, or manager training; they are the tier-2 individual-level supplement that NICE Guideline NG212 §1.6 treats as additive to those foundations, not a replacement for them.
Does it net out?
For anyone who wants the arithmetic rather than the assertion, there is a full worked example: a 50-employee firm taking weekly delivery, the programme cost net of tax, the reach, the break-even threshold, and a sensitivity analysis run across the inputs a sceptical reader is most likely to dispute. It has its own page, the cost model, which is also where the PSA gross-up, corporation-tax relief, and the P11D and co-pay alternatives live in full.
References
- Oxford Economics & Unum, The Cost of Brain Drain (2014). Average cost of replacing an employee earning £25,000 or more: £30,614 (2014 prices), decomposed as £5,433 logistical (advertising, agency fees, interview time, training) and £25,181 lost output during an average 28-week ramp-up to optimum productivity. Survey-based across five sectors including legal, accounting, IT, media and retail. UK CPI cumulative inflation 2014 to 2026 was approximately 41% (ONS series D7BT); on a like-for-like CPI basis the figure equates to roughly £43,200 in 2026. The original survey has not been re-run, so this is a price-level adjustment of the 2014 finding rather than a fresh measurement; sector-specific recruitment costs in law and finance may exceed the survey average. Partial modern triangulator on the recruitment-cost component: CIPD & Omni RMS, Resourcing and Talent Planning Report 2024 (Hogarth & McCartney, September 2024; mirror hosted on this site, SHA-256:
bfc3f885edb58de42e095e5cd7638c3875804565f58a3989030c2a649aa7e641; original) reports a 2024 median cost per hire of £2,000 for senior managers/directors (Table 1, p.13; n=246; private sector £2,000, public £2,500, non-profit £1,800), down from £3,000 in 2022 and £6,000 in 2017. The figure is broadly consistent with Oxford’s £5,433 “logistical” component once allowance is made for the median-vs-mean difference and downward pressure from recruitment-technology adoption. CIPD does not measure lost output during ramp-up, which was the larger Oxford component (£25,181 of the £30,614 total). No modern UK source has re-run a comparable lost-output methodology, so the headline current-price range continues to rest on the Oxford 2014 architecture with a CPI adjustment. A May 2026 review across the UK institutional research literature (CIPD, ONS, NIESR, IFS, Resolution Foundation, Bank of England Decision Maker Panel, the Skills and Employment Survey 2024, the Pissarides Review, Hays, Robert Walters, Cebr, and Oxford Economics’ and Unum’s own subsequent output) confirmed no post-2020 UK study has reproduced the lost-output-during-ramp construct, in £ or in weeks. The closest adjacent modern datapoint is Cebr × PageGroup × YouGov, “The cost of inefficient recruitment to UK businesses” (February 2025; n=100 UK organisations of 250+ staff, YouGov fieldwork, commissioned by a recruitment firm and therefore not assumption-free), estimating a £132.6 million aggregate UK productivity loss from inefficient recruitment and unfilled vacancies in 2025, equivalent to approximately nine working weeks of output per large organisation. That measures vacancy-side productivity gap, not the post-hire ramp curve Oxford modelled, and so reads as adjacent territory, not a substitute. The Oxford 2014 construct remains the most recent UK measurement of its kind at time of publication of this page; if a better-methodology UK source on this specific construct emerges, this footnote will be updated accordingly. - Pilot economics: 96 one-hour slots over the 12-week pilot (8 slots per weekly visit times 12 weeks). 30–50% slot uptake equals 29 to 48 sessions delivered. At one to two sessions per individual booker, the pilot reaches roughly 30 to 50 unique staff. Direct cost at £69 per session is £2,000 to £3,300; with the PSA gross-up at approximately 43% (basic rate), all-in £3,000 to £4,800.