Beyond “Ask for Angela”: What the Evidence Actually Shows About Keeping Bar Staff and Customers Safe

On 12 June 2026, ministers from four government departments sat down at 11 Downing Street with leaders from across hospitality, tourism and the night-time economy to discuss one thing: making venues safer, specifically for women and girls. Meetings like this usually produce warm words and little else. This one was different, because the head of the industry's own trade body used the moment to say something more pointed than the usual diplomatic language allows.

Kate Nicholls, CEO of UKHospitality, put it plainly:

"It's critical we work together to support our teams on the frontline, including the need to expand protections for retail staff to hospitality."

That's the industry's own leadership saying, on the record, that hospitality staff don't yet have something retail staff already have. We've written before about the wider staff-safety picture in licensed premises - the incidents, the legal duties, what good de-escalation training actually covers - in our existing guide to de-escalation training for bar staff. This piece asks a different question: what does the current push around guest safety actually add up to, and - more importantly - does any of it demonstrably work?

At a glance: the hospitality safety picture in 2026

The gap UKHospitality itself is naming

The roundtable wasn't a one-off. UKHospitality has an updated guest safety protocol currently out for consultation with the sector, covering room access procedures, protection of guest privacy, and the responsibilities of staff - developed with input from Rape Crisis and the Suzy Lamplugh Trust, and intended to sit alongside additional vulnerability training for frontline teams. It builds on a sector-specific guide UKHospitality already produces in partnership with the Equality and Human Rights Commission - a checklist and action plan for employers that the trade body itself describes as widely praised.

But look at where the sector actually stands today, in its own members' words. A UKHospitality member survey found 84% of respondents were already working proactively to prevent sexual harassment - a genuinely positive figure on its face. Look closer, though: only 46% had a pre-existing formal policy in place, with the remaining 38% relying on "some measures" short of that. Nearly four in ten hospitality businesses, by their own trade body's account, are doing something without doing it formally or consistently - which is precisely the gap between looking prepared and being prepared that runs through everything else in this piece.

Against that backdrop, Kate Nicholls' comment about extending retail-style protections to hospitality reads less like a soundbite and more like an honest admission: the sector knows its frontline teams are being asked to do more, without yet having either the legal parity or the consistent internal practice to match.

The problem with awareness schemes

Ask for Angela is the best-known initiative to come out of this space, and it deserves real scrutiny rather than automatic praise. The premise is simple: a customer who feels unsafe approaches staff and uses a codeword, and staff are trained to respond with a discreet intervention. It's a good idea. The evidence on how well it actually works in practice is uncomfortable.

A Serve Legal survey of more than 450 participants in May 2025 found that at venues actively displaying Ask for Angela branding, 76% of staff had received no training on how to respond. The Independent separately reported that roughly half of London pub staff surveyed couldn't recognise the codeword at all. The BBC's own reporting found staff who, when approached, simply had no idea what was being asked of them.

This isn't a scheme failing because the idea is wrong. It's failing because of something more specific and more fixable - and even the scheme's own people have said so:

"There is no copyright control over the advertising posters, allowing establishments to display support for the scheme without staff having received the necessary training - giving a false sense of security."

A poster is not a safeguard. It's a promise, and one that's currently being made without anything reliable behind it in a large proportion of venues. It's also worth naming a structural problem the scheme's director has acknowledged separately: hospitality's chronic staff turnover means a venue can train a cohort of staff thoroughly, then lose most of them within months - while the poster stays on the door long after the training has walked out of it. A safety scheme built on individual staff knowledge is only as strong as a sector's retention rate, and hospitality's retention rate is not strong.

Two duties, one training problem

Almost everything discussed so far is framed around protecting guests. There's a parallel legal picture that rarely gets connected to it: frontline and customer-facing teams themselves have their own, separate - and rapidly strengthening - protection against harassment by the people they serve.

Since October 2024, the Worker Protection Act 2023 has placed a proactive duty on employers to take reasonable steps to prevent sexual harassment of their staff. From October 2026, that changes significantly. The Employment Rights Act 2025 introduces employer liability for harassment of staff by third parties - customers, suppliers, delivery drivers, contractors, even other tenants sharing a building - with no incident threshold at all. Previously, a "three strikes" rule required two prior incidents before liability could attach; that threshold is gone. A single failure to take "all reasonable steps" is now enough, and law firm Mishcon de Reya has named hospitality specifically, alongside retail, as one of the sectors this will hit hardest, given how little control employers have over the third parties their staff interact with all day. It's also broader than most people assume: the duty covers harassment related to any protected characteristic - race, religion, sexual orientation - not sexual harassment alone. A breach carries a potential compensation uplift of up to 25% in tribunal proceedings.

Here's the point almost nobody is making explicitly: the training that satisfies a venue's duty to protect its own staff from customer harassment - recognising escalating or inappropriate behaviour, intervening safely, documenting what happened - is close to identical to the training needed to spot and respond to a vulnerable guest under the VAWG agenda. One sits under "guest safety." The other sits under employment law. In most venues, they're being planned, budgeted and delivered as two separate problems, when the actual skill underneath both is the same one.

It's worth noting what Mishcon itself recommends employers do to prepare: risk assessments, updated policies, stronger reporting channels - and staff training "refreshed periodically," specifically because a one-off session doesn't hold up as a defence if the same risk keeps recurring. That's a law firm independently arriving at the exact argument running through this entire piece.

Does training actually work? What the evidence says

It's worth being honest about this rather than just asserting that training obviously works. The Cochrane systematic review on de-escalation techniques found only two studies meeting its inclusion criteria worldwide - one in French dementia care homes - and concluded plainly that the reliability of the evidence was too low to say whether the techniques are effective. That's a genuinely humbling finding, and any provider who tells you the science is settled either hasn't read it or is choosing not to mention it.

The more useful evidence comes from applied research rather than pure trials. RAND Europe's evidence assessment for NHS Improvement found that de-escalation training may not, by itself, reduce the raw number of violent incidents - but it produced a significant reduction in lost workdays, improved staff retention, reduced complaints, and reduced overall expenditure. Crucially, where reductions in violence were found, they weren't always sustained through the study period - pointing to a need for periodic, refreshed training rather than a single session.

The single most relevant piece of evidence, though, comes from a NIOSH-funded study published in the Journal of Occupational and Environmental Medicine - and it directly included bars and restaurants in its sample, alongside convenience stores, liquor stores and motels. Researchers compared 305 businesses given training and a comprehensive security programme against 96 given nothing. The headline result was underwhelming: a 10% reduction in violent crime overall, not statistically significant. But when they split the intervention group by how well venues actually implemented the programme, the businesses with high compliance saw a real 26% reduction in violent crime - while low-compliance venues performed almost identically to the businesses that received no training at all.

That's the finding worth sitting with. It's not that conflict management training doesn't work. It's that most of the businesses studied didn't actually implement it properly - and the ones that did got a real result while the ones that didn't got nothing. The content wasn't the variable. The implementation was.

What this means for venues

Two things are true at the same time, and together they change the calculation for any venue operator. First, from October 2026, a single unaddressed incident of third-party harassment - no prior warning required - can expose an employer to liability, at a compensation uplift of up to 25%. Second, the clearest evidence available on conflict training says the deciding factor isn't whether a venue has a scheme or a policy document, but whether staff have actually, verifiably been trained and kept current.

Put those together and the honest conclusion isn't "buy a training course." It's that a poster, a policy sitting in a drawer, or a training day two years ago won't function as a defence under either the guest-safety agenda or the new legal duty - because neither is designed to reward good intentions. Both reward evidence of consistent, current practice. That's a harder thing to buy off the shelf than a laminated sign for the door, and a more honest thing to say than "our training reduces violence" - but it's what the research and the law are now both pointing toward.

Self-audit: is your venue's approach real, or just visible?

• Do all customer-facing staff - not just new starters - have a dated, auditable training record for both guest-safety and personal-safety scenarios?
• If your venue displays Ask for Angela or a similar scheme, could every member of staff on shift right now explain exactly what to do if approached?
• Is training refreshed on a set schedule, or did it happen once, at induction, and never again?
• Does your approach to protecting vulnerable guests and your approach to protecting staff from customer harassment sit in the same training, or two unconnected ones?
• If a new starter joined tomorrow, how long before they'd have the same level of preparedness as the rest of the team?

If more than one of these gives you pause, that's the gap between looking prepared and being prepared.

How Stand2's training closes the gap

The evidence above points to a specific conclusion, and it's worth restating precisely: in the one study that directly included bars and restaurants, high-compliance venues cut violent crime by 26%, while low-compliance venues - despite receiving the same training - saw no benefit at all. Implementation, not content, was the variable. That's exactly what Stand2's training is built to produce, rather than a one-off session that decays the moment turnover sets in.

All five of Stand2's training principles do real work here, not just the obvious ones. Ethical Decision-Making gives every member of staff a shared framework for recognising escalating or vulnerable behaviour and responding consistently, whether the person in front of them is a customer who needs de-escalating or a guest who needs protecting - the literal definition of the high-compliance consistency the evidence points to. Trauma-Informed & Context-Led practice means staff can tell the difference between someone becoming aggressive and someone becoming vulnerable, which is precisely the judgement the guest-safety agenda is asking of them. Prevention Before Intervention means more situations get defused before they reach the point of harassment, assault, or a reportable incident at all. Least Restrictive Practice keeps any necessary intervention proportionate and defensible - relevant to both a customer confrontation and a duty-of-care claim. And Safety First underpins all of it: staff who feel confident and supported are the ones who actually intervene, rather than staff running through motions they don't trust.

Every participant is entered onto the Stand2 Training Register with a uniquely numbered certificate - giving a venue a live, auditable answer to the question Ask for Angela's own critics have raised: not "do you display the poster," but "can you prove every member of staff on shift tonight is actually trained, and when they're due for a refresher." That's not a nice-to-have. It's precisely what Mishcon de Reya's own guidance recommends employers keep records of, and precisely the distinction the Casteel study found separated venues that got a real result from venues that didn't. We're currently developing a dedicated course pathway for hospitality and licensed premises; in the meantime, our existing de-escalation training for bar staff covers the frontline conflict management skills this piece has been discussing.

Where to start

If you want to understand what a properly implemented, tracked training programme could look like for your venue - one built to satisfy both the guest-safety conversation and your duty to protect staff from customer harassment, rather than treating them as separate problems - get in touch via our Course Finder.

Hospitality safety training: frequently asked questions

Does Ask for Angela actually work?

The evidence is mixed. A 2025 survey found 76% of staff at venues displaying the branding had received no training, and other reporting found around half of surveyed pub staff couldn't recognise the codeword. The scheme's own spokesperson has acknowledged there's no control over which venues display the posters without implementing real training behind them.

Are hospitality staff protected by the same law as retail workers?

Not currently in the same way. UKHospitality's own CEO has publicly called for retail worker protections to be extended to hospitality staff, indicating the sector doesn't yet consider itself to have equivalent protection.

What does the law require of hospitality employers on staff harassment?

Since October 2024, the Worker Protection Act 2023 has required employers to take reasonable steps to prevent sexual harassment of staff. From October 2026, the Employment Rights Act 2025 goes further, introducing employer liability for harassment of staff by third parties — including customers — with no prior-incident threshold, and covering all protected characteristics, not sexual harassment alone.

Does conflict management training actually reduce violence?

The evidence is more nuanced than most training providers suggest. A NIOSH-funded study of bars, restaurants and other retail-adjacent businesses found real reductions in violence — but only among venues with high compliance in implementing the training properly. Low-compliance venues saw no measurable benefit at all.

What's the difference between this and Stand2's existing bar staff training guidance?

Our existing guide covers frontline incidents, the legal duties on venue operators, and what good de-escalation training for bar staff should include. This piece focuses specifically on the guest-safety and vulnerability agenda, the evidence behind current safety schemes, and what the research says about whether training actually works.

Sources referenced

• GOV.UK, Government welcomes hospitality and tourism sector plans to further strengthen its safety standards to prevent violence against women and girls (12 June 2026)
• Acas, Protecting staff and customers in hospitality, Kate Nicholls, CEO of UKHospitality
• Mishcon de Reya, Harassment under the Employment Rights Act 2025: What are the key changes employers need to be aware of (30 June 2026)
• Serve Legal, 76% Untrained: Unpacking the 'Ask for Angela' Initiative
• The Independent, Fears as half of London pub staff fail to recognise Ask For Angela codeword
• BBC News, Ask for Angela: Staff had no clue, says pub worker
• The Verdict, 'Ask for Angela' — Should the Scheme be Legalised, or Should We Acknowledge its Inherent Flaws?
• Spencer S, Johnson P, Smith IC, De-escalation techniques for managing non-psychosis induced aggression in adults, Cochrane Database of Systematic Reviews, 2018
• RAND Europe, Reviewing the Evidence Base for De-escalation Training: A Rapid Evidence Assessment, prepared for NHS Improvement, 2019
• Casteel C, Chu LD, Greenland S, Kraus JF, Peek-Asa C, A Study of the Effectiveness of a Workplace Violence Intervention for Small Retail and Service Establishments, Journal of Occupational and Environmental Medicine, 2008, 50(12):1365-1370

This article reflects our understanding of current government initiatives, industry schemes, and the published research at the time of writing. It's intended as a practical briefing for venue operators and hospitality leaders, not legal advice.